Medical Intelligence Platform

One intelligence layer.
Every domain of modern medicine.

Clinical-grade AI that retrieves, reasons, and delivers structured intelligence across imaging, genomics, pathology, and clinical documentation.

Radiology·Oncology·Genomics·Diagnostics·Clinical Notes
Cortex Active
5 Domains Connected
33M+ Data Points
95.8%
Clinical reasoning accuracy
33M+
Structured data points
128k
Clinical pathways mapped
5
Medical domains covered
Intelligence Architecture

What the system processes.

Modern medicine requires simultaneous reasoning across disparate data types. AXIVIS is built to ingest and correlate the full clinical spectrum.

Medical Imaging
DICOM, CT, MRI, PET, US
Laboratory Data
Biomarkers, Pathology, Blood
Genomic Data
VCF, WES, WGS, RNA-seq
Procedural Media
Endoscopy, Video, Reports
Clinical Documentation
SOAP, Clinical Notes, EHR
One unified intelligence model. Continuous clinical reasoning.
Medical Intelligence

A clinician can only act on knowledge they can access in time.

Medical knowledge doubles every 73 days. No clinician can read it all. AXIVIS Medical Intelligence retrieves structured evidence from verified clinical literature at the point of care, and reasons across imaging, laboratory, genomic, and clinical data simultaneously through AXIVIS Cortex.

Medical knowledge doubles approximately every 73 days.
Clinicians miss actionable information in up to 75% of cases where relevant evidence exists but is not retrieved in time.
Explore Medical Intelligence
NEJM
LANCET
JAMA
Evidence Retrieved · Citation-Backed
CT
Imaging
Hb
Lab
VCF
Genomics
Unified Diagnostic Intelligence
Diagnostic Intelligence

Diagnostic errors are not always clinical failures. Often they are data failures.

Imaging interpreted without laboratory context. Laboratory results reviewed without imaging correlation. Procedural findings documented separately from genomic data. AXIVIS applies dedicated intelligence models to every diagnostic data type and surfaces findings within a single clinical environment.

Diagnostic errors affect 12 million US adults annually. Incomplete data is among the leading contributors.
40% of radiologists report workload levels that compromise diagnostic thoroughness.
Explore Diagnostic Intelligence
Oncology Intelligence

Cancer care generates more data than any other discipline in medicine. Most of it never connects.

Tumor imaging. Genomic mutations. Biomarker progression. Treatment history. Trial eligibility. Multidisciplinary decisions. AXIVIS consolidates these data streams into a single continuous intelligence model across the full cancer care pathway.

Diagnostic delays in oncology average 3 to 6 months from first symptom to confirmed diagnosis.
Treatment decisions made without integrated genomic data miss actionable targets in up to 30% of cases.
Explore Oncology Intelligence
<60s
Consultation to
completed record
Clinical Infrastructure

Medicine runs on documentation. It should not have to.

Physicians spend more time on records, orders, and administrative tasks than on patients. AXIVIS Clinical Infrastructure is the operational layer built to change that ratio.

49% of physician time spent on documentation, not care.
Fragmented records contribute to 12 million diagnostic errors annually.
Explore Clinical Infrastructure
Intelligence Architecture

Medical reasoning trained on clinical reality.

The intelligence layer underlying AXIVIS processes medical imaging, genomic data, laboratory results, electronic health records, and clinical documentation. Reasoning built for the data types that define modern clinical practice.

95.8%
Medical Reasoning Accuracy
Performance on MedQA clinical knowledge benchmark. Among the highest of any open model architecture at any scale.
33M+
Medical Image Training Pairs
Vision intelligence pre-trained on de-identified medical imaging data across radiology, histopathology, ophthalmology, and dermatology.
128k
Clinical Context Window
The model reasons across 128,000 tokens of clinical information simultaneously. Long-form patient records, imaging metadata, laboratory history, and genomic data processed together in a single session.
5
Medical Data Types
Medical imaging, genomic sequencing, laboratory results, electronic health records, and clinical documentation. All processed natively within the same intelligence layer without separate models per data type.
Benchmarks reflect the medical AI model architecture underlying the AXIVIS intelligence layer.
AXIVIS Black

Your complete biological picture. Analyzed. Not just recorded.

AXIVIS Black ingests your laboratory results, genetic data, imaging reports, and health records. AI analyzes them structurally, identifies risk patterns, tracks biological markers over time, and surfaces what requires attention. Not summaries. Not scores. Clinical-grade intelligence applied to your own data.

Most people have more health data than they can interpret. Most of it never connects.
Preventable chronic disease accounts for over 70% of global mortality. Most risk is detectable before symptoms appear.
HbA1c 5.2Glucose
BRCA1 WTGenomic
eGFR 92Renal
LDL 2.1Lipid

Advancing medical intelligence.
At the frontier of medicine.

Medical Intelligence

Clinical knowledge retrieved.
Patient data reasoned across.
Both at the point of care.

Medical knowledge relevant to a clinical decision exists in published literature, clinical guidelines, drug safety databases, and the patient's own history simultaneously. Accessing all of it manually, during a consultation, while managing the patient in front of you, is not possible.

Most clinical errors are not failures of competence. They are failures of access.
AXIVIS Medical Intelligence makes the full knowledge picture accessible at the point of care.
Deployment Options

Deploy Medical Intelligence standalone, or as part of the AXIVIS OS.

Standalone · AXV-01
Medical Intelligence on its own.
Procure Medical Intelligence as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Medical Intelligence as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Clinical Intelligence Search

Evidence retrieved from clinical literature in seconds. Citation-backed. At the point of care.

AXIVIS processes natural language medical queries against verified clinical literature, research publications, drug safety databases, and medical guidelines. The system retrieves relevant evidence, synthesizes it into a structured response, and presents citations linking to the original sources. Drug safety profiles, biomarker data, diagnostic considerations, and clinical research findings are all accessible through a single query. Voice input is supported. No literature search training required. No leaving the clinical workflow.

Example Workflow

A clinician encounters a rare drug interaction during a consultation. They submit a natural language query. AXIVIS retrieves relevant evidence from clinical literature, synthesizes a structured summary, and presents citations to the original sources. The answer is available before the consultation ends.

73 days
Medical Knowledge Doubling Time
The volume of published medical knowledge doubles approximately every 73 days.
75%
Of Actionable Information Missed
Clinicians identify clinical questions in most patient encounters but retrieve answers to fewer than half due to time and access constraints.
Citation-Backed Every Response
All synthesized responses include references linking to original source material. The clinician verifies, not trusts.
4

Input Types Processed Simultaneously

Medical imaging, laboratory results, genomic data, and clinical documentation analyzed together in a single session.

DICOM · PDF · VCF Supported Input Formats

CT, MRI, X-ray, DICOM imaging files, laboratory PDF reports, genomic VCF files, and clinical text documents all accepted within the same conversational session.

Diagnostic Considerations Not Diagnoses

AXIVIS Cortex surfaces clinical considerations for physician review. It does not diagnose. All outputs require clinician interpretation.

AXIVIS Cortex

Multimodal clinical reasoning across the full patient picture.

Clinical decision making rarely involves a single data type. A physician evaluating a complex patient is looking at imaging findings, laboratory results, genomic data, and clinical documentation simultaneously. AXIVIS Cortex processes all of these inputs together through a conversational interface. DICOM imaging studies, laboratory results, genomic data, and clinical documents are analyzed in combination. The system identifies relevant diagnostic considerations, suggests follow-up investigations, and generates structured analytical responses that reflect the complete patient picture. Not a single data point in isolation.

Axivis Cortex
Medical AI
Online
Imaging Analysis
DICOM findings, differentials
Oncology Protocols
Treatment lines, response criteria
Genomic Insights
Variants, targeted therapy
Lab Interpretation
Biomarkers, reference ranges
Literature Synthesis
Evidence, trials, guidelines
Drug Intelligence
Interactions, contraindications
›Summarise NCCN guidelines for stage IIIA NSCLC treatment.
›What are the imaging criteria for complete metabolic response on FDG PET/CT?
›Explain the clinical significance of EGFR exon 19 deletion in NSCLC.
›Review potential drug interactions between warfarin, amiodarone, and metformin.
Cortex Response
›Ask Cortex a clinical question... (Enter to send)
Send →
Session Context
0
Messages
0
Queries
Department Access
Radiology AI
Oncology Intelligence
Genomic Intelligence
Laboratory & Screening
Clinical Trials
Orders & Diagnostics
Query History
Processing...

A physician uploads a CT chest, a recent full blood count, and a genomic panel for a patient with an unresolved pulmonary finding. AXIVIS Cortex analyzes all three together, identifies a pattern consistent with an inflammatory process, surfaces relevant diagnostic considerations, and suggests two follow-up investigations. The physician reviews and proceeds.

How They Work Together

Clinical Intelligence Search

Retrieves and synthesizes knowledge from external medical literature. The input is a clinical question. The output is structured evidence from published sources with citations.

Used when a clinician needs to know what the literature says about a drug, a condition, a biomarker, or a clinical scenario.

+

AXIVIS Cortex

Reasons across the patient's own clinical data. The input is imaging, laboratory results, genomic reports, and clinical documents. The output is structured analytical reasoning about that specific patient's data.

Used when a clinician needs to understand what the patient's data means, not what the literature says in general.

Together they represent the full knowledge layer. External evidence and patient-specific reasoning. Both within the clinical workflow.
Who Medical Intelligence Serves

Physicians & Clinical Specialists

Evidence retrieval and multimodal patient data analysis supporting clinical decision making across specialties, from general practice through complex specialist care.

Diagnostic Staff

Structured analytical support for interpreting complex combinations of imaging, laboratory, and genomic data within the diagnostic workflow.

Clinical Researchers

Access to synthesized clinical literature, drug safety profiles, biomarker data, and research findings relevant to patient care and clinical research activities.

Medical Intelligence operates across the full platform.

Clinical Intelligence Search surfaces evidence relevant to diagnostic findings generated in Diagnostic Intelligence. AXIVIS Cortex reasons across imaging data from Radiology AI, laboratory trends from Labs & Screening, and genomic results from Genomic Intelligence. The knowledge layer connects to every domain.

Radiology AI
Oncology Intelligence
Clinical Infrastructure

Advancing medical intelligence.
At the frontier of medicine.

Oncology Intelligence

One continuous intelligence model across the full cancer care pathway.

An oncologist managing a single advanced cancer patient may review data from imaging, pathology, genomics, laboratory markers, and treatment records across multiple systems, teams, and time points. None of these systems were designed to talk to each other. The oncologist is the integration layer.

That is not a workflow problem. That is a clinical risk.
AXIVIS Oncology Intelligence is the integration layer.
Deployment Options

Deploy Oncology Intelligence standalone, or as part of the AXIVIS OS.

Standalone · AXV-02
Oncology Intelligence on its own.
Procure Oncology Intelligence as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Oncology Intelligence as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Oncology Intelligence · Live Product

A clinical operating environment for the full cancer care pathway.

Oncology Intelligence is deployed in active oncology departments today. The views below are from a live institutional instance — patient identifiers retained for illustration, interface unchanged. Every panel, status chip, and decision surface shown is the shipping product.

Oncology Critical Critical Lab · ANC 380 cells/μLS. Mitchell — ANC critically low. Immediate clinical review required for febrile neutropenia risk. → follow_up
Oncology High Disease Progression DetectedT. Brennan — RECIST assessment indicates progressive disease on latest CT. Treatment reassessment recommended. → radiology
Cases
5
Active oncology caseload
Critical
2
Immediate clinical review
On Treatment
3
Active regimens in flight
High Alerts
1
Progression · toxicity
Tox Alerts
4
CTCAE grade ≥2
Worklist Pending Orders Analytics Radiology Pathology Tumor Markers Staging Treatment Follow Up Tumor Board AlertsNEW MolecularNEW TrialsNEW
Oncology worklist with AI Clinical Assessment panel showing NSCLC staging and MDT referral
01 · Worklist

A caseload view that surfaces clinical priority before the oncologist opens a chart.

Every active patient is rendered as a clinical record card. Tumor type is tagged at the left (THORAC, MCRC, MEDIAS, PEMBRO). Status chips on the right indicate criticality (CRITICAL · URGENT · REVIEW), treatment cadence (ROUTINE · SCHEDULED), stage, and ECOG performance score. Active protocol and cycle are bound directly to each patient — Encorafenib 300mg OD + Cetuximab 500mg/m² Q2W, Cycle 4, Next: HOLD.

Toxicity events surface inline as grade-tagged chips: G4 Neutropaenia, G2 Hepatotoxicity, G2 Anaemia — aligned to CTCAE grading and routed into the department's tox-flag counters.

Patient CardMRN, age, sex, ONC-prefixed oncology case ID, presenting diagnosis, histology status, and clinical summary — every element structured, every element queryable.
Protocol ChipActive regimen with dosing, interval, cycle counter, and next-action decision (HOLD, CONTINUE, date-scheduled) bound to the patient record.
AI AssessmentRight-panel clinical synthesis runs in four tabs — AI Summary, Bloodwork, Toxicity, Timeline — with key clinical findings and attending notes visible beside the summary.
Guideline BindAI assessment references NICE NG122, ESMO protocols, and ECOG performance scoring by name — every recommendation cites the guideline that supports it.
Oncology radiology tab — interactive anatomical lesion distribution with disease status, tumor burden, tracked lesions, and AI decision support
02 · Radiology · RECIST

Anatomical lesion distribution with organ-level interactivity.

Radiology is not a list of reports. It is a continuous longitudinal model of disease. A patient-level anatomical canvas renders every tracked lesion in anatomical position, colour-coded by RECIST response (Responding, Stable, Progressing). Hovering any organ reveals site-specific findings — in this view, Small Intestine returns No findings; other regions surface lesion counts, SLD contribution, and prior-scan deltas.

Four department-grade KPIs frame the view: Disease Status (Awaiting Data · Responding · Stable · Progressing), Tumor Burden vs Baseline, Tracked Lesions with target / non-target split, and Reports Reviewed. Distribution filters segment the view across Organs · Lymph Nodes · Skeletal · Skin · Systemic.

Baseline AnchorFirst staging CT establishes baseline measurements. Every subsequent study auto-compared against baseline and prior timepoints using RECIST 1.1 response thresholds.
Lesion LedgerPer-lesion history — anatomical site, longest diameter, volumetric change, and response classification — persisted across every timepoint in the patient's longitudinal record.
Decision SupportAI-powered decision-support surface sits directly above the anatomical canvas — surfacing the recommended response classification and linking to the underlying imaging evidence.
Upload FlowNew radiology reports upload directly into the patient record. AI interpretation and lesion extraction run before the oncologist opens the study.
Oncology pending orders queue showing MDT, chemo, RT, and immunotherapy orders with detail panel and Take Order / Schedule / Assign / Upload / Reject actions
03 · Pending Orders

A single queue across MDT, chemotherapy, radiotherapy, and immunotherapy orders.

Oncology orders span modalities. A thoracic MDT referral for a suspected NSCLC, an Encorafenib/Cetuximab cycle review, an RT planning session for mediastinal lymphadenopathy, and a pembrolizumab maintenance cycle all flow through the same structured queue — each tagged by type (MDT · CHEMO · RT · IMMUNO), priority (URGENT · ROUTINE), and status (PENDING · REVIEW · SCHEDULED).

Selecting an order opens a structured detail panel on the right: Order ID, placing clinician, department, placed timestamp, due date, and clinical indication. Take Order · Schedule · Assign · Upload · Reject are the five institutional actions — every one logged, every one attributed.

Order TaxonomyMDT referrals, chemotherapy cycles, radiotherapy planning, immunotherapy maintenance, pathology review, imaging requests — all within a single unified queue with per-type filter chips.
Priority LogicIndication-driven priority assignment. Suspected NSCLC triggers MDT within 2 weeks per NICE NG122. Rising LFTs in the context of chemo trigger urgent CT and cycle hold.
Due TrackingEvery order carries a due date. Breaches flagged at the department level. Orders overdue by priority class surface at the top of the analytics critical-actions counter.
Audit ChainTake · Schedule · Assign · Upload · Reject — every action written to an immutable audit trail with clinician attribution and timestamp, aligned to medical device quality requirements.
Oncology Performance Dashboard — active cases, AI accuracy, protocol adherence, tox flags, critical actions, weekly throughput, treatment response distribution, oncologist workload
04 · Analytics

A departmental dashboard built for clinical governance, not vanity metrics.

The Oncology Performance Dashboard surfaces the five indicators that determine whether a cancer service is running to standard: Active Cases, AI Accuracy with override rate, Protocol Adherence against ESMO / NICE guidelines, Tox Flags with grade threshold, and Critical Actions requiring immediate review.

Weekly Case Throughput is segmented by treatment type (CHEMO · IMMUNO · MDT · RT). Treatment Response Distribution reports the full spread — CR · PR · SD · PD · NE — with a single institutional Disease Control Rate at the bottom. Oncologist Workload lists every attending with online status, active caseload, critical cases, tox alerts, and workload tier.

AI GovernanceAI Accuracy tracked against clinician override rate — every recommendation declined, refined, or accepted contributes to the governance loop. Override spikes trigger retraining review.
Protocol AdherenceESMO and NICE guideline adherence computed at the protocol level, not self-reported. Departures logged with rationale — clinician judgement, patient preference, or contraindication.
Response DistributionRECIST-based response classification aggregated across the department: Complete Response, Partial Response, Stable Disease, Progressive Disease, Not Evaluable. Disease Control Rate computed as CR + PR + SD / evaluable.
Workload BalancePer-oncologist active caseload, critical-case burden, and tox-alert count — visible to the clinical lead for workload balancing and on-call coverage planning.
Tumor Intelligence

Continuous lesion tracking from baseline to surveillance.

AXIVIS tracks tumor measurements across imaging studies using standardized oncology response assessment criteria. Baseline measurements are established at diagnosis. Every subsequent imaging study is compared against baseline. Change calculations, progression indicators, and response classifications are generated automatically and linked to the patient's longitudinal record. The oncologist reviews trends, not individual scans in isolation.

A follow-up CT is uploaded. AXIVIS compares it against the baseline and three prior studies, calculates change from prior, and response classification are ready before the radiologist begins manual review.

3–6 months
Average Diagnostic Delay
From first oncology symptom to confirmed diagnosis across major cancer types.

RECIST & iRECIST

Response Criteria

Standardized oncology response assessment frameworks applied automatically across treatment cycles.

Longitudinal

Measurement History

Every lesion tracked across every imaging study. Change calculations and progression trends visible in a single view.

Baseline
32mm
Cycle 2
26mm
Cycle 4
18mm
Cycle 6
10mm
Partial Response · RECIST 1.1 · -68.8% from baseline
PATHOGENIC
EGFR Exon 19 Del
Actionable · TKI
VUS
TP53 R273H
Monitoring
BENIGN
KRAS G12C
No action
3 Variants · 1 Actionable · Pharmacogenomic Cross-Referenced
Genomic Analysis

From raw sequencing data to actionable oncology intelligence.

Genomic findings are among the most complex data types in cancer care. AXIVIS ingests sequencing reports and VCF files, annotates variants for clinical significance, identifies pathogenic mutations relevant to treatment selection, and surfaces pharmacogenomic considerations that affect how the patient will respond to specific therapies. The oncologist receives a structured, prioritized genomic summary. Not a raw variant list.

30%
Of Cases Miss Actionable Targets
Treatment decisions made without integrated genomic data miss actionable therapeutic targets in up to 30% of advanced cancer cases.

Pathogenic · VUS · Benign

Variant Classification

Every identified variant classified by clinical significance and annotated with supporting literature references.

Pharmacogenomic

Drug-Gene Modelling

Variants affecting drug metabolism and treatment response identified and linked to current medication and treatment plans.

Biomarker Progression

Tumor marker trends that reflect treatment response over time.

Blood-based biomarkers are among the earliest indicators of treatment response and disease recurrence. AXIVIS records tumor marker measurements, generates longitudinal trend visualizations, and surfaces changes that may reflect disease activity or treatment efficacy. Reference ranges and trend indicators are displayed alongside historical values. The oncologist sees the trajectory, not just the latest number.

Longitudinal

Trend Visualization

Biomarker values plotted across the full treatment timeline. Change from prior visible at a glance.

Early Recurrence

Signal Detection

Rising biomarker trends flagged against established reference ranges and prior treatment response patterns.

Integrated

With Imaging Data

Biomarker trends correlated with tumor measurement data from the same treatment timeline.

4.2
Baseline
6.8
Cycle 2
5.1
Cycle 4
12.4
Cycle 6
PSA rising above reference range · Correlation with imaging recommended
Chemo
Carboplatin/Pemetrexed
Immuno
Pembrolizumab
Radio
54Gy/30#
Surgery
Lobectomy
4 Modalities · 12 Cycles · Active Treatment
Treatment Management

Every treatment decision documented. Every regimen tracked.

Cancer treatment involves multiple modalities, multiple cycles, and multiple decision points. AXIVIS documents treatment regimens across chemotherapy, immunotherapy, targeted therapy, radiation, and surgical interventions. Treatment timelines are visualized across the care pathway. Status is updated as therapy progresses. The complete treatment history is available in the same environment as imaging findings, genomic data, and biomarker trends.

Multi-Modality

Treatment Documentation

Chemotherapy, immunotherapy, targeted therapy, radiation, and surgical records in one timeline.

Cycle-Level

Tracking

Individual treatment cycles documented with status, dates, and clinical notes at each stage.

Connected

To Imaging & Genomics

Treatment response correlated with tumor measurements and biomarker progression across the same timeline.

Clinical Trial Matching

Eligibility identified from molecular and clinical criteria automatically.

Clinical trial eligibility is determined by a combination of cancer type, stage, genomic profile, treatment history, and demographic factors. Manually reviewing trial eligibility for every patient against every available study is not feasible in a high-volume oncology practice. AXIVIS evaluates the patient's clinical and molecular profile against available trials and surfaces potential matches for oncologist review. The oncologist makes the eligibility determination. AXIVIS ensures no trial is missed because the data was never compared.

A patient with refractory non-small cell lung cancer has an EGFR exon 20 insertion identified on genomic analysis. AXIVIS surfaces three trials with eligibility criteria that match the patient's molecular profile and treatment history.

40,000+
Patients Identified For Trials
Across precision oncology networks using AI-assisted molecular matching.

Molecular

& Clinical Criteria

Trial matches based on genomic variants, cancer type, stage, treatment history, and demographic eligibility factors.

Oncologist

Makes Final Determination

Matching is presented as informational support. Eligibility confirmation and enrollment remain with the clinical team.

94% Match
KEYNOTE-789 · Phase III
Pembrolizumab + Pemetrexed · EGFR Exon 20 · Recruiting
87% Match
CHRYSALIS-2 · Phase I/II
Amivantamab · EGFR/MET · Active, Not Recruiting
72% Match
PAPILLON · Phase III
Amivantamab + Chemo · EGFR Exon 20 Ins · Enrolling
Tumor Board · Case #2847
Onc
Rad
Path
Gen
Surg
Nrs
Recording · Consensus Recommendations · Linked to Patient Record
Multidisciplinary Care

Tumor board decisions documented. Care team coordination structured.

Cancer care decisions involving multiple specialties require a shared record of what was discussed, who participated, and what was recommended. AXIVIS provides structured documentation for multidisciplinary tumor board meetings including participating specialists, clinical data reviewed, and consensus recommendations. Surveillance activities scheduled after treatment are tracked against the patient record. Follow-up tasks are visible to the care team. Nothing is left to informal communication.

MDT Decisions

Fully Documented

Meeting details, participating specialists, data reviewed, and consensus recommendations recorded and linked to the patient record.

Surveillance

Scheduling

Post-treatment follow-up activities scheduled, tracked, and flagged when overdue against the patient's care timeline.

Care Team

Visibility

Every member of the oncology care team sees the same patient record, the same treatment history, the same open tasks.

Who Oncology Intelligence Serves

Oncologists & Oncology Nurses

Medical, surgical, and radiation oncologists and their nursing teams managing the full cancer care pathway from diagnosis through long-term surveillance.

Cancer Centers & Oncology Departments

Institutional oncology programs requiring structured documentation, multidisciplinary coordination, and integrated data management across the care team.

Precision Medicine & Genomics Teams

Clinical genetics specialists and precision medicine teams translating complex genomic and molecular data into actionable treatment decisions.

Advancing medical intelligence.
At the frontier of medicine.

AXV-10 · Procedure Intelligence

AI-assisted procedural video analysis.
Frame-level. Real-time. Clinician-validated.

Deep learning analysis of endoscopic, surgical, and interventional procedural video. Frame-by-frame detection of polyps, mucosal abnormalities, bleeding sites, and anatomical landmarks. Structured procedural reporting with complication tracking across the full intervention lifecycle.

EU MDR Class IIa · Frame-Level CNN Analysis · MP4/MOV Ingestion · Human-in-the-Loop Mandatory
Deployment Options

Deploy Procedure Intelligence standalone, or as part of the AXIVIS OS.

Standalone · AXV-10
Procedure Intelligence on its own.
Procure Procedure Intelligence as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Procedure Intelligence as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Procedure Intelligence · Live Product

A complete procedural workstation, reading room to report signature.

Procedure Intelligence is deployed in active endoscopy, bronchoscopy, and gynaecology suites today. The following walks the full cycle — departmental worklist, per-finding report drill-down, and serial procedure comparison — all from a live institutional instance. Patient identifiers retained for institutional realism; interface unchanged.

Critical Lab · ENDO · AI 96% S. Mitchell · MRN-447821Hb 6.8 g/dL — critical anaemia · Plt 18 ×10⁹/L — critical thrombocytopaenia · Pancytopaenia. Cross-flagged to endoscopic procedure — bleeding risk elevated. → acknowledge · dismiss
Procedures Worklist Pending Procedures 3 Analytics
Procedures Today
4
Currently reporting
Critical / STAT
0
Immediate review
Awaiting Review
2
Scientist queue
AI Accuracy
97.4%
Flag PPV
Avg TAT
34m
vs 2hr manual
Types Active
3
ENDO · BRONCH · GYNAE
Procedure Intelligence worklist — four procedureFindings across endoscopy, bronchoscopy, and gynaecology with priority, AI findings, and status chips
01 · Procedures Worklist

A departmental worklist that triages itself.

The worklist opens to the full departmental procedureFindings queue. Six KPIs anchor the view: Procedures Today, Critical/STAT, Awaiting Review, AI Accuracy with flag PPV, Avg TAT with manual baseline, and Types Active across the procedural domains in flight. Today's queue shows four active procedureFindings across three domains — two endoscopy, one bronchoscopy, one gynaecology.

Every row carries the full institutional procedural ledger: Order No (PO-2026-xxxxx) with corresponding procedure accession (PROC-xxxxx), priority chip (URGENT · ROUTINE · STAT), patient identity with MRN and age, procedure type chip with test composition, AI-extracted preliminary finding with confidence score, status chip, scientist attribution, timestamp, and View Report · View Results actions.

Triage SortDefault sort is clinical priority — URGENT · REVIEW surfaces above ROUTINE · NORMAL. Scientists see the patient who needs them next, not the patient who was scoped first.
Procedure TaxonomyChip-coded procedural classification — ENDO · BRONCH · GYNAE · CYSTO · LAP · ENT · HYST. Each chip colour-coded, each domain filterable across the full queue.
AI FindingsAI-extracted preliminary finding rendered in report-ready prose with per-finding confidence score — 82% · 12% · 74% · 8% across today's four cases. Three-tier status chip (CRITICAL · REVIEW · NORMAL) drives sort and colour.
Cross-Module FlagCritical values from Labs & Screening cross-flag into Procedure Intelligence for the same patient. A pancytopaenia flag surfaces as a bleeding-risk advisory on any active endoscopic procedure.
Procedure Report · Live Context

Every procedure opens into a structured procedural report.

Selecting any row from the worklist opens the corresponding procedural report. The report carries the full patient chrome bar, a seven-tab navigation surface, and an AI SCORE badge at the top right — in this case 82% · REVIEW for Sarah Mitchell's urgent gastroscopy. Two institutional actions at the top of every report: + Addendum and Sign Report.

SM
Sarah Mitchell
DOB: 1968-04-12 · Age: 57 · MRN-447821
Order
PO-2026-04301
Accession
PROC-94201
Study Date
2026-03-30
Referring
Dr. S. Kapoor
Scientist
Dr. R. Harwood
Panel
Gastroscopy (OGD) + Biopsy
AI Score
82%
REVIEW
Interpretation Media / AI Key Findings Impression Structured Report Prior Procedure Metadata
Procedure Report Key Findings tab showing Irregular Z-line with Prague classification and Hiatus Hernia with prior-current comparison
02 · Key Findings · Procedure Report

Procedural findings as clinical record cards.

The Key Findings tab renders every finding as a structured card — procedural title, interval-change chip (STABLE · WORSENING · IMPROVING · NEW), severity status (REVIEW · NORMAL · CRITICAL), proc type, ICD-10 code, clinical description in the specialist vocabulary, AI confidence bar, and a dense PRIOR → CURRENT comparison row.

First card: Irregular Z-line — Suspected Barrett's Oesophagus · STABLE · REVIEW · Proc Type ENDO · ICD-10 K22.7 · 84% confidence. Description: "Irregular Z-line at 38cm from incisors. Salmon-coloured mucosa extending 2–3cm circumferentially proximal to GEJ. Prague classification C2M3. Three targeted biopsies taken. No dysplasia seen macroscopically." Prior: "C1M2 — biopsies negative for IM". Current: "C2M3 — biopsies pending".

Second card: Hiatus Hernia · STABLE · NORMAL · ENDO · ICD-10 K44.9 · 96% confidence. Description: "Sliding hiatus hernia approximately 3cm. GEJ at 38cm, diaphragmatic impression at 41cm. No strangulation or obstruction identified." Prior: "2cm hiatus hernia". Current: "3cm hiatus hernia".

Prague ClassificationBarrett's oesophagus findings carry the international Prague C&M classification (C2M3) inline in the clinical description — the specialist reads without cross-referencing.
ICD-10 BindingEvery procedural finding tagged with the corresponding ICD-10 code — K22.7 Barrett's oesophagus · K44.9 Diaphragmatic hernia — queryable for cohort analysis and billing without free-text parsing.
Prior → CurrentEvery finding carries an inline delta — C1M2 → C2M3, 2cm → 3cm, negative for IM → biopsies pending. Interval change is the finding, not a footnote.
Confidence LayerPer-finding AI confidence bar — 84% for the Z-line · 96% for the hernia. Reasoned discrimination: confidence lower for the finding awaiting histology; high for the structural measurement.
Procedure Report Prior Procedure tab with two-year timeline, Prior vs Current cards, finding comparison, and prior-current impressions side-by-side
03 · Prior Procedure · Serial Surveillance

Procedural surveillance rendered as a two-procedure timeline.

The Prior Procedure tab renders the full surveillance arc for this patient. A PROCEDURE TIMELINE anchors the view with the prior procedure on the left and the current procedure on the right, separated by the interval elapsed — in this patient's case, 2.0 years (748 days).

Below the timeline, a PRIOR PROCEDURE vs CURRENT PROCEDURE comparison renders the structured ledger for both procedures side-by-side — Date, Operator, Instrument, and AI disposition. Then the FINDING COMPARISON ledger: every finding from both procedures aligned with a stability chip, interval-change text, and confidence bar. Finally, PRIOR IMPRESSION vs CURRENT IMPRESSION side-by-side prose — the reporting clinician sees both reports in the same frame, with two years of elapsed surveillance in between.

Procedure TimelineTwo-procedure timeline with interval days rendered inline — 2.0 years (748 days). Prior and current procedure cards anchored at each end with institutional metadata.
Operator ContinuityOperator and instrument tracked across procedures — in this patient, Dr. R. Harwood on the Olympus GIF-H290 gastroscope for both timepoints. Continuity of operator is a documented quality indicator.
AI Disposition DeltaAI disposition at both timepoints rendered side-by-side — NORMAL 14% (prior) → REVIEW 82% (current). The shift is the clinical signal the reporting clinician acts on.
Impression DiffPrior and current impressions rendered as adjacent prose cards — the diff is visible in a single glance rather than buried in prior-study retrieval.
04 · Timeline · Rendered Live

The surveillance arc, rendered in the page.

The same two-procedure timeline from the product is reconstructed below in the page — an active interval-flow showing the two years of surveillance elapsed between prior gastroscopy and current gastroscopy for this patient, with Prior vs Current procedure cards, finding comparison ledger, and impression diff rendered live.

Procedure Timeline · Sarah Mitchell · 2.0 years surveillance arc
◷
Gastroscopy (OGD) + Biopsies
2024-03-12
PROC-88341
2.0 years (748 days)
◷
Gastroscopy (OGD) + Biopsy
2026-03-30
PROC-94201
Prior Procedure
Date2024-03-12
OperatorDr. R. Harwood
InstrumentOlympus GIF-H290 gastroscope
AI NORMAL 14%
Current Procedure
Date2026-03-30
OperatorDr. R. Harwood
InstrumentOlympus GIF-H290 gastroscope
AI REVIEW 82%
Finding Comparison
2 current · 2 prior
→
stable
Irregular Z-line — Suspected Barrett's Oesophagus REVIEW K22.7
2 years ago: C1M2 — biopsies negative for IM → C2M3 — biopsies pending
84%
Details ⌄
→
stable
Hiatus Hernia NORMAL K44.9
2 years ago: 2cm hiatus hernia → 3cm hiatus hernia
96%
Details ⌄
Prior Impression · 2024-03-12
Irregular Z-line at 38cm. Prague C1M2. Biopsies ×4 — histology: no intestinal metaplasia, no dysplasia. Hiatus hernia 2cm. No active lesions.
Current Impression · 2026-03-30
Irregular Z-line with Prague C2M3 pattern at GEJ. Hiatus hernia 3cm. Biopsies taken. Histology awaited to confirm or exclude Barrett's oesophagus with intestinal metaplasia. No active bleeding. Antrum and duodenum normal.
05 · Pending Procedures

Pre-procedural orchestration across the full scheduling pipeline.

The Pending Procedures surface handles the full procedural order lifecycle — from clinician request through consent, pre-procedural optimisation, list scheduling, and theatre assignment. Three pending orders active on the current session:

ENDO
Fatima Al-Rashid · 68F · MRN-990142 · OGD + H. pylori Test
Gastroenterology · Dr. E. Mbeki · PO-2026-04320 · Indication: epigastric pain, weight loss, iron deficiency
STAT PENDING
BRONCH
Priya Nair · 71F · MRN-776203 · Flexible Bronchoscopy + EBUS-TBNA
Respiratory Medicine · Dr. S. Osei · PO-2026-04318 · Indication: hilar lymphadenopathy, staging confirmation
URGENT SCHEDULED
GYNAE
Anna Kowalski · 38F · MRN-443092 · Diagnostic Hysteroscopy
Gynaecology · Dr. K. Mbeki · PO-2026-04311 · Indication: post-menopausal bleeding, endometrial thickening on USS
ROUTINE PENDING
Clinical IndicationEvery order carries a structured clinical indication written by the referring clinician — "epigastric pain, weight loss, iron deficiency", "hilar lymphadenopathy, staging confirmation", "post-menopausal bleeding, endometrial thickening on USS" — rendered in the referral vocabulary.
Department RoutingOrders labelled with originating department — Gastroenterology · Respiratory Medicine · Gynaecology — and ordering clinician, for workload distribution and list scheduling.
Row ActionsTake Order · Schedule · Assign · Consent · Upload · Reject — six institutional actions per pending row. Each action attributed, logged, and audit-traced to the procedural lifecycle.
Pre-Procedural OptimisationCross-module flagging surfaces pre-procedural concerns from Labs & Screening — anticoagulation status, critical bloods, allergy documentation — before the patient enters the procedure suite.
06 · Analytics

A departmental dashboard built for procedural governance.

The Analytics surface renders the quality indicators that determine whether a procedural service is running to standard — throughput, turnaround, detection, quality, and clinician workload. Six quality indicators anchor the procedural governance ledger:

Polyp Detection Rate
96.4%
Caecal Intubation Rate
94.2%
Adenoma Detection Rate
38.1%
Avg Withdrawal Time
7m 48s
Complication Rate
0.4%
Clinician Override
3/142
Quality IndicatorsEndoscopy quality indicators computed from frame-level analysis — Adenoma Detection Rate · Caecal Intubation Rate · Withdrawal Time · Bowel Preparation Quality. Computed from the video, not self-reported.
Complication TrackingPost-procedural complications — bleeding, perforation, sedation-related events, post-polypectomy syndrome — tracked against the procedural ledger for departmental review.
Per-Endoscopist AuditPer-clinician quality indicator breakdown — ADR, CIR, withdrawal time, complication rate — rendered at the individual endoscopist level for JAG / BSG audit requirements.
Override LoopAI override rate computed from every declined or refined recommendation. 3 overrides out of 142 reviewed — spikes trigger a retraining review flag for the governance loop.
Supported Procedures

Five procedural domains. One intelligence layer.

Each procedure type processed by dedicated deep learning models trained on annotated procedural video datasets.

EGD
Upper GI Endoscopy
Oesophageal, gastric, and duodenal mucosal assessment with lesion characterisation
COL
Colonoscopy
Polyp detection, Paris classification, pit pattern analysis, withdrawal quality
BRN
Bronchoscopy
Airway inspection, endobronchial lesion detection, vocal cord assessment
CYS
Cystoscopy
Bladder mucosal mapping, tumour detection, haematuria source localisation
LAP
Laparoscopy
Surgical field analysis, instrument tracking, anatomical landmark recognition
Frame-Level Video Analysis

Every frame analysed. Every finding annotated. Nothing missed in transit.

Procedural video is ingested as MP4/MOV and processed frame-by-frame through modality-specific convolutional neural networks. The system identifies anatomical landmarks, detects abnormalities in real time, characterises findings by morphology and location, and generates timestamped annotations throughout the procedure. Unlike static imaging, procedural video requires continuous temporal analysis — a finding visible for 0.3 seconds during withdrawal is as clinically significant as one visible for 30 seconds.

Continuous Temporal Analysis

30 FPS Processing · Temporal Correlation

Each frame analysed independently and correlated temporally. Findings tracked across sequential frames to distinguish transient artefacts from persistent abnormalities.

Polyp Detection & Characterisation

Paris Classification · Pit Pattern · Size Estimation

Detected polyps classified by Paris morphology (0-Ip, 0-Is, 0-IIa, 0-IIb, 0-IIc, 0-III), Kudo pit pattern analysis, and AI-estimated size. Each detection includes confidence score and anatomical segment location.

Mucosal Abnormality Detection

Inflammation · Erosion · Ulceration · Vascular Pattern

Beyond polyps, the system detects mucosal inflammation, erosive changes, ulceration, abnormal vascular patterns, and early neoplastic changes across the assessed mucosal surface.

Procedure Recording · Live Analysis
REC
00:00
12:45
AI Detections · Current Procedure
Sessile Polyp — Ascending Colon
Paris 0-Is · Kudo III-L · Est. 8mm · Frame 4,291
0.94
Pedunculated Polyp — Sigmoid
Paris 0-Ip · Kudo IV · Est. 12mm · Frame 8,107
0.91
Flat Lesion — Transverse Colon
Paris 0-IIa · Kudo II · Est. 5mm · Frame 11,433
0.82
Mucosal Erythema — Caecum
Non-specific inflammation · Biopsy recommended · Frame 2,018
0.76
Diverticulum — Sigmoid
Anatomical landmark · No abnormality · Frame 7,845
0.88
Real-Time Detection Panel

Findings detected, classified, and logged as they appear.

As procedural video is analysed, detected findings populate a real-time detection panel visible to the endoscopist during the procedure. Each finding includes classification (polyp type, mucosal abnormality, anatomical landmark), morphological characterisation, estimated size, anatomical segment location, frame reference, and confidence score. The detection panel serves as a second observer — augmenting the endoscopist's visual assessment without interrupting procedural flow.

Confidence-Scored Detections

Per-Finding Certainty Assessment

Every detection includes a confidence score (0.0–1.0). High-confidence findings (above 0.85) flagged with critical markers. Lower-confidence findings presented for clinician verification. Threshold configurable per institution.

Frame-Reference Linking

Finding → Frame → Timestamp

Each detection linked to the exact video frame where it was identified. Clinicians click any finding to jump directly to the relevant frame for visual review and confirmation.

A colonoscopy procedure is recorded. AXIVIS processes the video at 30 frames per second. At frame 4,291 (02:23), a sessile polyp is detected in the ascending colon with 94% confidence. The endoscopist sees the detection marker appear on the live feed and confirms the finding for removal. Post-procedure, the structured report includes all detections with frame references, timestamps, and classifications.
Procedure Timeline Intelligence

The complete procedure reconstructed. Phase by phase.

AXIVIS automatically segments the procedural video into anatomical phases — insertion, navigation, examination, intervention, and withdrawal. Each phase is timed, measured, and annotated. For colonoscopy, withdrawal time is calculated automatically against minimum quality standards (6-minute threshold). Adenoma detection rate is tracked per clinician over time. The system transforms procedural video from a passive recording into a structured, searchable, auditable clinical dataset.

Automated Phase Segmentation

Insertion · Navigation · Examination · Intervention · Withdrawal

AI identifies procedural phases based on visual landmarks, instrument positioning, and anatomical transitions. Phase boundaries marked automatically with timestamp references.

Quality Metrics

Withdrawal Time · ADR · Caecal Intubation · Bowel Prep Score

Key procedural quality indicators calculated automatically from video analysis. Institutional benchmarking against published quality standards. Per-clinician longitudinal quality tracking.

Procedure Phase Timeline
00:00
INSERT
Insertion
01:54
NAVIGATE
Navigation
05:02
EXAMINE
Examination
09:28
INTERVENE
Intervention
10:58
WITHDRAW
Withdrawal
Total: 12:45
Withdrawal: 7:43 ✓
Findings: 5
Caecal: Confirmed
Structured Procedural Report
AI-Generated · Pending Review
Procedure Summary
Colonoscopy — Diagnostic with Polypectomy. Indication: Surveillance following prior adenoma resection. Sedation: Propofol. Preparation: Boston Bowel Prep Scale 7/9 (L:2 T:3 R:2). Caecal intubation confirmed by appendiceal orifice and ileocaecal valve visualisation. Total procedure time: 12 minutes 45 seconds. Withdrawal time: 7 minutes 43 seconds (above 6-minute quality threshold).
Findings
Finding 1: Sessile polyp, ascending colon, 8mm, Paris 0-Is, Kudo III-L pattern. Removed by cold snare polypectomy. Sent for histology. Finding 2: Pedunculated polyp, sigmoid colon, 12mm, Paris 0-Ip, Kudo IV pattern. Removed by hot snare with submucosal injection. Haemostasis confirmed. Finding 3: Flat lesion, transverse colon, 5mm, Paris 0-IIa. Biopsied. Finding 4: Non-specific mucosal erythema, caecum. Biopsied. Finding 5: Diverticulum, sigmoid — documented, no intervention.
Complications
None. No perforation. No uncontrolled bleeding. Post-polypectomy sites haemostatic. Patient tolerated procedure well.
Structured Procedural Reporting

Complete procedural reports generated. Clinician edits before finalisation.

AXIVIS generates structured procedural reports that include procedure summary (indication, sedation, preparation quality), all findings with frame references and AI classification, interventions performed, complications assessment, and follow-up recommendations. The report integrates AI detections, clinician confirmations, and procedural quality metrics into a single structured document. The performing clinician reviews, edits, and approves before the report enters the patient record.

Automated Quality Scoring

BBPS · Withdrawal Time · ADR · Caecal Intubation

Quality metrics extracted automatically from video analysis and integrated into the procedural report. Institutional quality dashboards track metrics per clinician over time.

Complication Documentation

Perforation · Bleeding · Cardiopulmonary · Infection

Complications documented with severity grading, timing, management actions, and outcome. Structured complication reporting enables institutional safety monitoring and regulatory compliance.

Bleeding Detection & Haemostasis Monitoring

Active bleeding detected. Haemostasis status tracked continuously.

During interventional procedures, AXIVIS continuously monitors the visual field for signs of active bleeding — arterial spurting, oozing, submucosal haematoma formation. When bleeding is detected, the system logs the onset frame, anatomical location, estimated severity, and tracks the visual field continuously until haemostasis is confirmed. Post-intervention bleeding surveillance continues through the withdrawal phase. The bleeding detection module provides an additional safety layer for procedures involving polypectomy, biopsy, or mucosal resection.

Severity Grading

Arterial · Venous · Oozing · Controlled

Detected bleeding classified by visual characteristics. Arterial spurting flagged as critical with immediate alert. Oozing monitored with sustained tracking until resolution.

Haemostasis Confirmation

Post-Intervention Surveillance

After intervention, the system monitors the polypectomy or biopsy site through subsequent frames. Haemostasis confirmed or flagged if bleeding recurs. Time-to-haemostasis logged in the procedural report.

Bleeding Event Timeline
Onset
09:31
Post-polypectomy
Intervention
09:34
Clip applied
Haemostasis
09:36
Confirmed
Time to haemostasis: 5 min 12 sec · Severity: Venous oozing · Status: RESOLVED
Clinical Safeguards

Every detection advisory. Every finding clinician-confirmed.

Advisory Output

All AI detections are decision support. The performing endoscopist or surgeon retains full clinical authority over procedural decisions. No autonomous intervention.

Clinician Confirmation

Each AI detection requires clinician acknowledgement — confirmed, modified, or dismissed. Only confirmed findings enter the structured procedural report.

Model Traceability

Every frame-level inference logged with model version, input frame reference, detection output, confidence score, and the clinician who reviewed the finding.

Video Integrity

Original procedural video preserved without modification. AI annotations are overlaid, never embedded. Full procedural recording available for retrospective review.

5
Procedure Types
Colonoscopy, upper GI endoscopy, bronchoscopy, cystoscopy, and laparoscopy — all processed natively.
30fps
Frame Analysis Rate
Every frame of procedural video analysed independently with temporal correlation across sequential frames.
<200ms
Detection Latency
From frame capture to AI detection overlay. Sub-200ms latency enables real-time procedural augmentation without delay.
4
Quality Metrics
Withdrawal time, adenoma detection rate, caecal intubation, and bowel preparation quality — all calculated automatically.
Who Procedure Intelligence Serves

Endoscopists

Gastroenterologists performing diagnostic and therapeutic endoscopy using AI-augmented detection to maximise adenoma detection rate and procedural quality compliance.

Surgeons

Surgical teams using laparoscopic and minimally invasive approaches with AI-assisted anatomical landmark recognition and procedural phase documentation.

Endoscopy Departments

Department leadership with visibility into procedural volumes, quality metrics, complication rates, and per-clinician performance benchmarking across the unit.

Advancing medical intelligence.
At the frontier of medicine.

AXV-07 · Orders & Diagnostics

Every diagnostic order tracked.
From request to result. Nothing lost in transit.

State-based diagnostic order lifecycle management across every department. Unique order identifiers, real-time status tracking, clinician notifications, and immutable audit trail. Orders transition through defined states — Pending, In Progress, Completed, Cancelled — with enforced governance at every stage. Completed orders become read-only. No diagnostic leakage.

EU MDR Class IIa · State-Based Lifecycle · Immutable Audit · Cross-Department Orchestration
Deployment Options

Deploy Orders & Diagnostics standalone, or as part of the AXIVIS OS.

Standalone · AXV-07
Orders & Diagnostics on its own.
Procure Orders & Diagnostics as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Orders & Diagnostics as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Order Lifecycle

Four states. One direction. Full governance.

Every diagnostic order follows a defined state machine. Transitions are enforced, logged, and irreversible. No order can skip a state or revert without explicit audit documentation.

Ordered
Clinician creates order with patient, test type, priority, clinical indication
Pending
Order queued for processing. Department notified. Priority assigned.
In Progress
Sample collected or imaging scheduled. Processing underway.
Completed
Results available. Order becomes read-only. Clinician notified.
Cross-Department Order Management

Every order visible. Every department connected. Every result delivered.

Diagnostic orders span departments — a physician orders blood tests processed by the laboratory, imaging studies interpreted by radiology, histology specimens analysed by pathology, and genetic testing performed by the genomics unit. AXIVIS provides a unified order management layer where every order is visible regardless of which department processes it. The ordering clinician sees real-time status updates. The processing department receives structured order details. Results flow back to the patient record automatically upon completion.

Unique Order Identifiers

ORD-YYYY-XXXXX Format

Every order receives a system-generated unique identifier. No duplicate orders. No ambiguous references. Each identifier links to the ordering clinician, patient, test type, clinical indication, priority, and complete audit trail.

Priority Classification

STAT · Urgent · Routine · Scheduled

Orders classified by clinical urgency. STAT orders trigger immediate department notification. Priority-based queue sorting ensures the most clinically critical orders are processed first.

Clinician Notifications

Real-Time Status Updates

Ordering clinicians receive notifications at each state transition — order acknowledged, sample collected, processing started, results available. Critical results trigger immediate push notification to mobile.

Active Order Queue · All Departments
Order ID
Description
Department
Ordered
Status
ORD-2026-08421
CT Head — Acute stroke query
Radiology
14:23
STAT
ORD-2026-08420
FBC, U&E, CRP, LFT, Coagulation
Laboratory
14:18
IN PROG
ORD-2026-08419
Tissue biopsy — Sigmoid polyp
Pathology
13:45
IN PROG
ORD-2026-08417
BRCA1/2 Panel — Hereditary risk
Genomics
11:30
PENDING
ORD-2026-08415
Chest X-Ray — Pre-operative
Radiology
10:52
DONE
ORD-2026-08412
HbA1c, Lipid Panel, TSH
Laboratory
09:15
DONE
ORD-2026-08408
MRI Brain — Follow-up
Radiology
Yesterday
CANCELLED
Diagnostic Leakage Monitor
Orders created (this month)
1,247
Results received
1,198
Results reviewed by ordering clinician
1,184
Results pending review >48 hours
14
Orders without results >7 days
3
Leakage rate
0.24%
3 orders flagged for follow-up escalation
Diagnostic Leakage Prevention

An ordered test with an unreviewed result is a patient safety failure.

Diagnostic leakage — the failure to follow up on ordered diagnostic tests — is among the most common and preventable causes of delayed diagnosis. Studies estimate that 7–8% of abnormal results are never communicated to patients. AXIVIS tracks every order from creation through result delivery to clinician review. Orders that produce results but remain unreviewed trigger escalating notifications. Orders that fail to produce results within expected timeframes are flagged for investigation. The system ensures that no diagnostic result falls into the gap between departments.

Result Review Tracking

Ordered → Resulted → Reviewed → Actioned

Each order tracked not just to result delivery, but to clinician review and documented action. A result that arrives but is never reviewed is functionally equivalent to a result that never arrived.

Escalation Protocol

24h → 48h → 72h → Department Head

Unreviewed results trigger automated escalation. Initial reminder at 24 hours. Follow-up at 48 hours. At 72 hours, the order is escalated to the department head with audit documentation.

Institutional Dashboard

Leakage Rate · Review Compliance · Turnaround Time

Leadership dashboard showing institutional diagnostic leakage rate, per-department review compliance, mean turnaround time by order type, and flagged orders requiring follow-up.

Immutable Audit Trail

Every action logged. Every transition timestamped. Every actor identified.

Every diagnostic order maintains a complete, immutable audit trail from creation through completion. Each state transition records the timestamp, the clinician or system that initiated the transition, the reason for the transition, and the order state before and after. Completed orders become read-only — no modification, no deletion, no retroactive changes. The audit trail is available for clinical governance review, regulatory inspection, and medico-legal documentation at any time.

State Transition Logging

Timestamp · Actor · Reason · Before/After State

Every transition between order states captured with full metadata. System-initiated transitions (e.g., auto-escalation) documented with the triggering rule and governance policy reference.

Read-Only Completion

Immutable After Finalisation

Once an order reaches Completed state and results are attached, the order record becomes immutable. No fields can be modified. Amendments require a new linked order with explicit cross-reference to the original.

A CT Head is ordered as STAT at 14:23 for an acute stroke query. The order transitions to Pending at 14:23, acknowledged by Radiology at 14:25, In Progress at 14:31 (patient on table), and Completed at 14:48 with the structured report attached. The ordering clinician reviews the result at 14:52. Total order lifecycle: 29 minutes. Every transition logged with clinician identity and timestamp.
Audit Trail · ORD-2026-08421
14:52
Result Reviewed
Dr. K. Okonkwo · ED Consultant
14:48
Completed · Report Attached
Dr. R. Harwood · Reporting Radiologist
14:31
In Progress · Patient on Table
CT Suite 2 · Radiographer: M. Chen
14:25
Acknowledged by Department
Radiology · Auto-prioritised: STAT
14:23
Order Created · STAT Priority
Dr. K. Okonkwo · Indication: Acute stroke query, onset 45 min
Clinical Safeguards

Every order governed. Every result tracked to review.

State Enforcement

Orders follow a defined state machine. No state can be skipped, reverted, or bypassed. Each transition requires authenticated clinician action or documented system trigger.

Leakage Prevention

Every order tracked from creation through result review. Unreviewed results trigger escalating alerts. Orders without results beyond expected timeframes flagged automatically.

Immutable Records

Completed orders become read-only. No modification, no deletion. Amendments require a new linked order with explicit audit cross-reference to the original.

Cross-Department Visibility

Every order visible to the ordering clinician regardless of which department processes it. No orders lost in inter-departmental handoffs.

4
Order States
Pending, In Progress, Completed, and Cancelled — enforced state machine with no bypass or reversion.
4
Priority Levels
STAT, Urgent, Routine, and Scheduled — with department-specific notification rules and queue sorting.
100%
Audit Coverage
Every state transition, every actor, every timestamp logged immutably from order creation to result review.
<1%
Target Leakage Rate
Institutional diagnostic leakage rate target — with automated escalation for unreviewed results beyond 48 hours.
Who Orders & Diagnostics Serves

Ordering Clinicians

Physicians and specialists creating diagnostic orders with real-time visibility into order status, result delivery, and processing timeline across every department.

Processing Departments

Laboratory, radiology, pathology, and genomics departments receiving structured order details with priority classification and clinical indication for efficient processing.

Clinical Governance

Medical directors and clinical governance teams monitoring diagnostic leakage rates, review compliance, turnaround times, and order lifecycle performance across the institution.

Advancing medical intelligence.
At the frontier of medicine.

AXV-08 · Labs & Screening

Laboratory intelligence.
Extracted. Classified. Contextualised.

Upload lab reports in any format — PDF, image, HL7. AI extracts every biomarker, classifies abnormal values against reference ranges, generates contextual clinical interpretation, and tracks longitudinal trends across sequential reports. Integrated clinical screening tools for mental health, cervical cytology, and dermatological assessment.

EU MDR Class IIa · Multi-Format Ingestion · Longitudinal Trending · Clinician Review Mandatory
Deployment Options

Deploy Labs & Screening standalone, or as part of the AXIVIS OS.

Standalone · AXV-08
Labs & Screening on its own.
Procure Labs & Screening as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Labs & Screening as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Laboratory AI · Live Product

An institutional results workstation, not a PDF viewer.

Laboratory AI is deployed in active laboratory operations today. The six screens below trace a single critical haematology panel for Sarah Mitchell — 57-year-old patient, STAT HAEM order, Hb 6.8 g/dL, critical thrombocytopaenia, pancytopaenia — through every surface the clinical scientist and treating clinician interact with: departmental worklist, per-finding drill-down, structured prior comparison, full report interpretation, pending request orchestration, and departmental analytics.

Critical Lab Value · HAEM · AI 96% S. Mitchell · MRN-447821 · LAB-83401Hb 6.8 g/dL — critical anaemia. Plt 18 ×10⁹/L — critical thrombocytopaenia. Pancytopaenia. → acknowledge · dismiss
SM
Sarah Mitchell
DOB: 1968-04-12 · Age: 57 · MRN-447821
Order
LO-2026-04221
Accession
LAB-83401
Study Date
2026-03-30
Referring
Dr. S. Kapoor
Scientist
Dr. A. Okafor
Panel
Full Blood Count + Differential + ESR
Interpretation Key Findings Impression Structured Report Prior Results Metadata
Laboratory AI results worklist with six KPI cards and four result rows showing STAT, URGENT, ROUTINE priorities
01 · Results Worklist

The departmental results list that triages itself.

The Laboratory AI worklist opens to the full departmental results queue. Six KPIs anchor the view: Results Today, Critical/STAT, Awaiting Review, AI Accuracy, Avg TAT with comparison against manual baseline, and Panels Active.

Results Today
4
Currently reporting
Critical / STAT
1
Immediate review
Awaiting Review
3
Scientist queue
AI Accuracy
97.4%
Flag PPV
Avg TAT
34m
vs 2hr manual
Panels Active
4
HAEM · BIOCHEM · TUMOUR · MICRO

Every result row carries the full institutional ledger: accession number, priority chip (STAT · URGENT · ROUTINE), patient identity, panel taxonomy chip (HAEM · BIOCHEM · TUMOUR · MICRO), scientist attribution, and the AI-generated preliminary finding with confidence score. The first row — Sarah Mitchell · HAEM · 96% · CRITICAL: Hb 6.8 g/dL severe anaemia, WBC 0.9 critical leucopaenia, Plt 18 critical thrombocytopaenia — surfaces at the top by triage sort, not time.

Triage SortDefault sort is clinical priority, not chronological. CRITICAL surfaces above REVIEW surfaces above NORMAL. Scientists see the patient who needs them next, not the patient who arrived first.
Panel TaxonomyChip-coded panel classification — HAEM (haematology), BIOCHEM (biochemistry), TUMOUR (tumour markers), MICRO (microbiology), COAG, IMMUNO, HORMONE, SCREEN. Filterable across the full queue.
AI FindingsAI-extracted preliminary finding rendered in report-ready prose with per-finding confidence score. Three-tier status chip (CRITICAL · REVIEW · NORMAL) drives sort and colour.
Scientist AttributionEvery result attributed to the responsible clinical scientist with laboratory affiliation and timestamp — Dr. A. Okafor · AXIVIS General · 2026-03-30 07:45.
Prior Result Comparison with 6 WORSENING 2 STABLE ribbon and per-analyte comparison cards
02 · Prior Results

Serial laboratory change rendered as clinical record cards.

Opening a report with prior results available surfaces the Prior Results tab automatically. An interval-change ribbon declares the departmental posture up front: 6 WORSENING · 2 STABLE. Three view modes — Cards, Chart, Table — for the same underlying data.

Each analyte renders as a structured card: gauge-indicator of current value against reference range, critical status chip, reference range, current → prior arrow with absolute and percentage change. Haemoglobin: 6.8 g/dL · Critical Low · Ref: 12–16 · 8.2 → 6.8 · ▼ 1.4 (-17.1%). Four structured fields — Current · Prior · Reference · Trend — followed by a Position Within Reference Range bar rendering the analyte's position as a marker on the normal band. Critical values carry an inline banner: "Critical value — immediate clinical review required. Notify treating clinician."

Interval RibbonDepartment-level posture surfaced at the top of the comparison: WORSENING · STABLE · IMPROVING counts. Scientist sees the trajectory of the full panel before any single analyte.
Reference PositionPer-analyte position-within-reference-range bar renders the current value against the normal band — visual indication of how far outside reference the analyte sits.
View ModesCards for clinical review cadence, Chart for longitudinal trending across multiple prior timepoints, Table for dense numeric comparison — same data, three lenses.
Critical BannerCritical-value inline banner with explicit clinical action — "immediate clinical review required · notify treating clinician". Acknowledgement routes to the audit trail.
Key Findings drill-down for Haemoglobin Severe Anaemia with AI detection confidence, result values parameters, descriptors, clinical significance
03 · Key Findings · Drill-Down

Drill into a finding and see the full analyte panel.

Opening any abnormal result opens a dedicated finding page. Title, panel chip, severity status, and ICD-10 coding at the top — in this view: Haemoglobin — Severe Anaemia · HAEM · ICD-10: D64.9 — Anaemia, unspecified · CRITICAL. AI Detection Confidence renders as a proportional bar at 96%.

RESULT VALUES & PARAMETERS surfaces the complete diagnostic vocabulary: Haemoglobin 6.8 g/dL (ref 12–16), MCV 104 fL (ref 80–100), MCH 35.1 pg (ref 27–33), Reticulocytes 0.4% (ref 0.5–2.5), ESR 88 mm/hr (ref <20) — every analyte with its own reference range rendered inline.

DESCRIPTORS renders the haematological pattern in report-ready prose: "Hb 6.8 g/dL — severely below reference range (12–16 g/dL). Macrocytic pattern (MCV 104 fL). Reticulocytopenia (0.4%) indicating impaired bone marrow response rather than haemolytic or haemorrhagic aetiology." CLINICAL SIGNIFICANCE translates pattern into action: "This result requires urgent clinical attention. Immediate correlation with clinical presentation and relevant specialist review is recommended."

ICD-10 BindingEvery finding tagged with the corresponding ICD-10 code — D64.9 Anaemia, unspecified — queryable for cohort analysis, billing, and epidemiological review without free-text parsing.
Pattern RecognitionDescriptors block renders the haematological pattern — macrocytic, microcytic, normocytic, reticulocytopenia, reticulocytosis — in the clinical vocabulary the haematologist uses, not generic numeric prose.
Reference BindingEach result value rendered with its own reference range inline. 104 fL (ref 80–100). Scientist sees the deviation without cross-referencing a separate document.
Clinical ActionClinical Significance block translates the pattern into a clinical posture — "urgent clinical attention · immediate correlation · specialist review". The scientist has the action, not just the finding.
Interpretation tab with test panels status list, laboratory interpretation prose, recommendation block, and prior results comparison table
04 · Interpretation

A full laboratory report written in the cadence of a haematologist.

The Interpretation tab opens with TEST PANELS — an ordered list of every analyte against its severity chip: Haemoglobin · Critical Low. White Blood Cells · Critical Low. Platelets · Critical Low. Neutrophils · Critical Low. Lymphocytes · Low. MCV · High. ESR · High. Reticulocytes · Low.

LABORATORY INTERPRETATION renders the full clinical synthesis in the prose conventions used by reporting haematologists: "Severe pancytopaenia with macrocytosis. Bone marrow failure pattern in context of known myeloid malignancy. Neutrophils critically low — infection risk extreme."

RECOMMENDATION surfaces explicit clinical actions: "Urgent haematology review. Consider G-CSF. Blood cultures if febrile. Transfusion threshold reached for both RBC (Hb <7) and platelets (<20)." PRIOR RESULTS AVAILABLE FOR COMPARISON renders every analyte as Prior → Now with a directional arrow. Eight analytes. Six trending down (▼▼). Two trending up (▲▲ for MCV and ESR — both clinically consistent with the interpretation).

Test Panel ListOrdered analyte list with per-analyte status chip — Critical Low · Low · Normal · High · Critical High. Scientist sees the full panel posture before reading prose.
Clinical ProseLaboratory Interpretation rendered in the specialist vocabulary actually used in reports: pancytopaenia · macrocytosis · bone marrow failure pattern · myeloid malignancy — not generic numeric summary.
RecommendationExplicit clinical recommendation — haematology review · G-CSF · blood cultures · transfusion thresholds — with threshold values (Hb <7, Plt <20) named inline.
Trend DirectionalPrior → Now arrows (▼▼ double-down for worsening, ▲▲ for rising) rendered beside each analyte — the trajectory visible at a glance.
Pending Requests queue with 8 orders, filter chips, priority chips, and order detail panel with Take Order Schedule Assign Upload Reject actions
05 · Pending Requests

The laboratory order queue in one department-level view.

The Pending Requests surface handles the full laboratory order lifecycle. Two filter rows at the top: status (ALL · STAT · URGENT · ROUTINE · PENDING · IN_PROGRESS · COMPLETED · REJECTED) and panel (HAEM · BIOCHEM · MICRO · COAG · IMMUNO · TUMOUR · SCREEN).

HAEM BIOCHEM MICRO COAG IMMUNO TUMOUR HORMONE SCREEN

Every request row carries panel chip, patient identity, priority chip, order ID, test panel composition, originating department, ordering clinician, and specimen type. Selecting a row opens a structured detail panel on the right: Order ID LO-2026-04290, Clinician Dr. E. Mbeki, Department Oncology, Specimen EDTA, Placed 2026-03-30 06:00, Due 2026-03-30 09:00, Indication: "Pre-cycle bloods. Osimertinib cycle 8. Neutropenia monitoring." Five row actions — Take Order · Schedule · Assign · Upload · Reject.

Test CompositionEvery order carries full test composition — FBC + Differential + Reticulocytes, LFTs + U&Es + Magnesium + TSH, PT + APTT + Fibrinogen + D-Dimer, CEA + CA 19-9 + AFP + LDH, PSA + Free PSA + PSA Density — rendered in the test catalogue the scientist already knows.
Specimen BindingEvery order labelled with required specimen — EDTA · SST Serum · Citrated Plasma · Questionnaire. Scientist sees specimen requirements before taking the order.
Clinical IndicationEvery order carries a full clinical indication — "Pre-cycle bloods. Osimertinib cycle 8. Neutropenia monitoring". Scientist reviews intent, not just panel composition.
Row ActionsTake Order · Schedule · Assign · Upload · Reject — five institutional actions per pending row. Each action attributed, logged, and audit-traced.
Laboratory Analytics dashboard with six KPIs, hourly throughput, results by panel, scientist workload, and AI interpretation performance metrics
06 · Analytics

A departmental dashboard built for laboratory governance.

Six KPIs anchor the Laboratory Analytics surface: Results Today 186, STAT Pending 6, Avg TAT 34m, Critical 8, Reported 142, Awaiting 44. A stacked HOURLY RESULT THROUGHPUT chart runs from 06:00 through 17:00.

RESULTS BY PANEL renders the departmental distribution as horizontal bars with percentages: HAEM 42 (23%) · BIOCHEM 58 (31%) · TUMOUR 24 (13%) · MICRO 19 (10%) · COAG 14 (8%) · IMMUNO 11 (6%) · SCREEN 18 (10%). SCIENTIST WORKLOAD lists every clinical scientist — Dr. Okafor 48 tests / 3 critical / 28m TAT / ONLINE · Dr. Adeyemi 39 tests / 1 critical / 32m · Dr. Mbeki 31 tests / 0 critical / 41m / OFFLINE · Dr. Osei 24 tests / 2 critical / 38m.

AI INTERPRETATION PERFORMANCE surfaces the six governance metrics that matter:

Critical Detection
98.1%
Flag Accuracy
97.4%
Interp PPV
93.2%
False Neg Rate
0.8%
Results Reviewed
186
Clinician Override
4/186
Critical DetectionSensitivity of the AI to genuine critical values — 98.1%. Computed against the scientist's final classification as ground truth.
False Neg RateGenuine critical values missed by AI interpretation — 0.8%. The single most important laboratory governance metric. Surfaced at the top of the performance ledger.
Interp PPVPositive predictive value of the full laboratory interpretation — the probability that an AI-generated interpretation is clinically confirmed by the reporting scientist. 93.2%.
Override TrackingEvery AI interpretation declined or refined by the scientist contributes to the governance loop. 4 overrides out of 186 reviewed — spikes trigger a retraining review flag.
Supported Input Formats

Three ingestion pathways. One structured output.

Lab reports arrive in every format. AXIVIS normalises them all into structured, queryable clinical data.

PDF
Document Reports
PDF lab reports from any laboratory. OCR extraction with biomarker identification and value parsing
IMG
Image Capture
Photographed or scanned lab results. Vision AI extracts structured data from printed reports
HL7
Electronic Feed
HL7 v2 and FHIR R4 structured laboratory data. Direct integration with laboratory information systems
Biomarker Extraction & Classification

Every value extracted. Every abnormality flagged. Every trend tracked.

Uploaded laboratory reports are processed through multi-stage extraction. AI identifies individual biomarkers, parses numeric values and units, maps each result against age- and sex-adjusted reference ranges, and classifies findings as Normal, Abnormal, or Critical. The system generates contextual interpretation for each result — not just whether a value is high or low, but what the clinical significance may be in the context of the patient's history, medication list, and concurrent findings.

Intelligent Value Parsing

OCR · NLP · Unit Normalisation

Biomarker names, values, units, and reference ranges extracted from unstructured documents using combined OCR and natural language processing. Units normalised to standardised formats for cross-laboratory comparability.

Reference Range Classification

Age-Adjusted · Sex-Adjusted · Pregnancy-Aware

Each result classified against demographically appropriate reference ranges. Paediatric, geriatric, and pregnancy-specific ranges applied automatically based on patient profile.

Contextual Interpretation

Clinical Significance · Differential Considerations

Beyond flag-and-reference classification, AI generates brief contextual interpretation for abnormal findings — potential clinical significance, common differential considerations, and suggested follow-up investigations.

Extracted Biomarker Panel · Complete Blood Count
Marker
Value
Reference
Interpretation
Status
Haemoglobin
9.8 g/dL
12.0–16.0
Below reference — evaluate for iron deficiency, chronic disease, or haemolysis
LOW
WBC
14.2 ×10⁹/L
4.0–11.0
Elevated — consider infection, inflammation, stress response, or haematological cause
HIGH
Platelets
245 ×10⁹/L
150–400
Within reference range — no action required
OK
Creatinine
142 µmol/L
60–110
Elevated — assess renal function with eGFR, hydration status, and nephrotoxic medications
CRIT
eGFR
48 mL/min
>60
Stage 3a CKD — monitor trend, renal referral if sustained decline
CRIT
HbA1c
7.2%
<6.5%
Above diabetic target — review glycaemic control, medication adherence, dietary factors
HIGH
TSH
2.4 mIU/L
0.4–4.0
Within reference range — euthyroid
OK
CRP
28 mg/L
<5
Significantly elevated — active inflammation or infection, correlate clinically
CRIT
Longitudinal Biomarker TrendHbA1c · 12 months
Trend: Rising +0.6% over 12 months
Target: <6.5%
Longitudinal Biomarker Intelligence

A single result is a snapshot. A trend is a trajectory.

Individual lab results provide limited clinical value in isolation. The same value that is normal for one patient may represent a significant deterioration for another when viewed against their baseline. AXIVIS automatically matches current results against all prior laboratory data for the same patient, calculates the trajectory, and alerts clinicians to clinically significant trends — rising creatinine suggesting progressive renal impairment, falling haemoglobin indicating ongoing blood loss, or climbing HbA1c reflecting deteriorating glycaemic control.

Automated Trend Detection

Slope Analysis · Rate of Change · Trajectory Alerts

Each biomarker tracked longitudinally with automated trend calculation. Sustained directional changes flagged with rate of change and projected trajectory. Alert thresholds configurable per biomarker and per institution.

Cross-Biomarker Correlation

Multi-Analyte Pattern Recognition

Simultaneous trends across related biomarkers identified automatically — rising creatinine with falling eGFR confirms renal trajectory; falling haemoglobin with rising ferritin suggests inflammatory anaemia rather than iron deficiency.

A patient presents with fatigue. Current haemoglobin is 11.2 g/dL — within the low-normal range. AXIVIS retrieves prior results: 13.1 six months ago, 12.4 three months ago. The trend is a sustained 1.9 g/dL decline over six months. The system flags this as a clinically significant downward trajectory and recommends investigation for occult blood loss, even though the current value alone would not trigger an alert.
Integrated Clinical Screening

Validated screening instruments. Structured scoring. Longitudinal tracking.

Standardised clinical screening tools integrated natively within the platform. Scores calculated automatically, severity classified, and results tracked over time.

PHQ-9
Depression Screening
Patient Health Questionnaire-9. Validated severity scoring (0–27) with classification: Minimal, Mild, Moderate, Moderately Severe, Severe. Longitudinal mood trajectory tracking.
GAD-7
Anxiety Assessment
Generalised Anxiety Disorder 7-item scale. Severity classification (Minimal, Mild, Moderate, Severe) with treatment threshold identification and response monitoring.
MoCA
Cognitive Screening
Montreal Cognitive Assessment. Multi-domain evaluation: visuospatial, naming, attention, language, abstraction, delayed recall, orientation. Scored 0–30 with normative comparison.
CRV
Cervical Screening
Cervical cytology and HPV result processing. Bethesda classification (NILM, ASC-US, LSIL, HSIL, AGC). Colposcopy referral pathway integration.
DRM
Dermatology Analysis
Skin lesion image analysis with dermoscopic pattern recognition. ABCDE criteria assessment, lesion evolution tracking, and referral pathway for suspicious findings.
CVD
Cardiovascular Risk
Integrated cardiovascular risk calculation using laboratory data (lipid panel, HbA1c, CRP) with clinical parameters (BP, BMI, smoking status, family history). QRISK3 and Framingham scoring.
Clinical Safeguards

Every extraction verified. Every interpretation clinician-reviewed.

Extraction Verification

AI-extracted values displayed alongside the source document for visual verification. Clinicians confirm extracted values match the original report before data enters the patient record.

Advisory Interpretation

All contextual interpretations are decision support. Clinical significance assessment requires qualified clinician judgement in the context of the full patient presentation.

Critical Value Alerting

Results classified as critical trigger immediate notification to the ordering clinician. Critical alert acknowledged receipt logged with timestamp and clinician identity.

Source Document Retention

Original laboratory documents (PDF, image, HL7) retained alongside extracted structured data. Full provenance from source document to structured record maintained.

3
Input Formats
PDF documents, photographed images, and HL7 electronic feeds — all normalised into structured biomarker data.
6
Screening Tools
PHQ-9, GAD-7, MoCA, cervical screening, dermatology analysis, and cardiovascular risk — all with longitudinal tracking.
<30s
Report Processing
From document upload through OCR extraction, biomarker identification, classification, and contextual interpretation generation.
3
Classification Tiers
Normal, Abnormal, and Critical — with age-adjusted, sex-adjusted, and pregnancy-aware reference range matching.
Who Labs & Screening Serves

Physicians & GPs

Primary care and specialist physicians reviewing laboratory results with AI-augmented interpretation, trend analysis, and cross-biomarker correlation at the point of clinical review.

Laboratory Directors

Laboratory leadership with visibility into result turnaround, critical value notification compliance, and extraction accuracy metrics across the laboratory operation.

Clinical Screening Teams

Mental health, cervical screening, and preventive health teams using integrated validated screening instruments with automated scoring and longitudinal outcome tracking.

Advancing medical intelligence.
At the frontier of medicine.

AXV-12 · Genomic Intelligence

Native genomic data processing.
From sequencing file to clinical decision.

Direct VCF ingestion with automated variant annotation, ACMG classification, pharmacogenomic cross-referencing, and clinical actionability scoring. Germline and somatic analysis unified within a single clinical intelligence environment. Every genomic output advisory — clinician interpretation mandatory.

EU MDR Class IIa · VCF Native · ACMG/AMP Classification · Germline + Somatic
Deployment Options

Deploy Genomic Intelligence standalone, or as part of the AXIVIS OS.

Standalone · AXV-04
Genomic Intelligence on its own.
Procure Genomic Intelligence as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Genomic Intelligence as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Supported Sequencing Inputs

Four sequencing modalities. One genomic intelligence layer.

Every major clinical sequencing format processed natively through the same variant interpretation pipeline.

WGS
Whole Genome
Complete genomic coverage. Germline risk, structural variants, non-coding regions
WES
Whole Exome
Protein-coding regions. Cost-effective clinical-grade variant detection
TGP
Targeted Panel
Gene-specific panels for hereditary cancer, cardiac, pharmacogenomic screening
PGx
Pharmacogenomics
Drug metabolism profiling. CYP450, DPYD, TPMT, UGT1A1 genotyping
Variant Classification Engine

Every variant annotated. Every classification evidence-based.

Uploaded VCF files are processed through a multi-stage annotation pipeline. Each variant is cross-referenced against ClinVar, gnomAD, COSMIC, and institutional databases, then classified using ACMG/AMP criteria into five tiers: Pathogenic, Likely Pathogenic, Variant of Uncertain Significance (VUS), Likely Benign, and Benign. Classification evidence is documented — population frequency, functional predictions, segregation data, and literature references. The clinical geneticist reviews every classification before it enters the patient record.

ACMG/AMP Five-Tier Classification

Pathogenic · Likely Pathogenic · VUS · Likely Benign · Benign

Standardised classification using the American College of Medical Genetics criteria. Each tier determined by aggregation of evidence codes (PVS1, PS1-4, PM1-6, PP1-5, BA1, BS1-4, BP1-7).

Multi-Database Cross-Reference

ClinVar · gnomAD · COSMIC · OMIM · PharmGKB

Each variant queried against population frequency databases, clinical significance repositories, somatic mutation catalogues, and pharmacogenomic knowledge bases simultaneously.

Clinical Actionability Scoring

Actionable · Monitor · No Current Action

Beyond classification, each pathogenic or likely pathogenic variant assessed for clinical actionability — does this finding change screening, treatment, or management for this patient?

Variant Classification Panel
12 variants · 3 actionable
BRCA1 c.5266dupC
PATHOGENIC
Exon 20 · Frameshift · p.Gln1756Profs*74 · gnomAD: 0.0001 · ClinVar: Pathogenic (reviewed)
ACTIONABLE — Enhanced screening + risk-reducing surgery discussion
TP53 R175H
LIKELY PATHOGENIC
Exon 5 · Missense · Somatic · VAF 34% · COSMIC: Hotspot · Functional: Loss-of-function
ACTIONABLE — Therapeutic target assessment + trial eligibility
CHEK2 I157T
VUS
Exon 3 · Missense · Germline · gnomAD: 0.004 · ClinVar: Conflicting · Functional: Uncertain
MONITOR — Reclassification watch + family segregation recommended
APC I1307K
LIKELY BENIGN
Exon 16 · Missense · Germline · gnomAD: 0.02 (Ashkenazi) · ClinVar: Likely benign
NO CURRENT ACTION — Population-specific polymorphism
MLH1 -93G>A
BENIGN
Promoter · Germline · gnomAD: 0.32 · ClinVar: Benign · No functional impact
NO CURRENT ACTION — Common polymorphism
Germline + Somatic Analysis

Hereditary risk and tumour-specific mutations. Analysed together.

Clinical genomics requires distinguishing between inherited germline variants (present in every cell, passed to offspring, informing hereditary risk) and acquired somatic mutations (tumour-specific, informing treatment selection). AXIVIS processes both analysis types through the same pipeline, with clear labelling and distinct clinical pathways. Germline findings trigger hereditary risk assessment and family screening recommendations. Somatic findings feed into AXV-11 Oncology Intelligence for treatment pathway integration.

Germline Analysis

Hereditary Risk · Carrier Status · Family Implications

Hereditary cancer syndromes (BRCA1/2, Lynch, Li-Fraumeni), cardiac conditions (LQTS, HCM), pharmacogenomic profiles. Findings trigger cascade testing recommendations for at-risk family members.

Somatic Analysis

Tumour Profiling · Therapeutic Targets · Resistance Mechanisms

Tumour-specific variant allele frequencies, driver mutations, actionable therapeutic targets, resistance mutations. Integrated with oncology pathway for treatment decision support.

Oncology Integration

AXV-12 → AXV-11 Cross-Module Intelligence

Genomic findings from AXV-12 feed directly into AXV-11 Oncology Intelligence. Variant classifications, therapeutic targets, and trial eligibility criteria are available within the oncology treatment pathway without data re-entry.

Pharmacogenomic Intelligence

The right drug. The right dose. Determined by genotype, not trial and error.

Pharmacogenomic profiling identifies how a patient metabolises specific drugs based on their genetic profile. AXIVIS cross-references genotyped pharmacogenes (CYP2D6, CYP2C19, CYP3A4, DPYD, TPMT, UGT1A1) against the prescribed medication list to generate dosing guidance. Poor metabolisers, intermediate metabolisers, and ultra-rapid metabolisers identified with clinical recommendations. This intelligence integrates with AXV-04 Consultation Suite prescription safety checking — genomic drug-gene interactions flagged at the point of prescribing.

Drug-Gene Interaction Mapping

Genotype → Metaboliser Status → Dosing Guidance

Each pharmacogene genotyped and mapped to metaboliser phenotype. Dosing recommendations generated per CPIC and DPWG guidelines. Interactions flagged before prescribing occurs.

Prescription Integration

AXV-12 → AXV-04 Cross-Module Safety

Pharmacogenomic findings automatically available to the Consultation Suite prescription safety engine. When a clinician prescribes a drug with a known pharmacogenomic interaction, the system flags the genotype-specific risk before the prescription is finalised.

Pharmacogenomic Profile · Drug-Gene Interactions
Clopidogrel
CYP2C19
Poor metaboliser (*2/*2) — reduced active metabolite formation
ADJUST
Codeine
CYP2D6
Ultra-rapid metaboliser (*1/*1xN) — increased morphine conversion risk
AVOID
5-Fluorouracil
DPYD
Intermediate metaboliser (*1/*2A) — 50% dose reduction required
ADJUST
Warfarin
CYP2C9
Normal metaboliser (*1/*1) — standard dosing appropriate
STANDARD
Simvastatin
SLCO1B1
Normal function (521TT) — standard myopathy risk
STANDARD
Tamoxifen
CYP2D6
Extensive metaboliser (*1/*1) — enhanced endoxifen conversion
ENHANCED
Clinical Safeguards

Every classification advisory. Every variant clinician-interpreted.

Advisory Classification

All AI-generated variant classifications are decision support. Final interpretation remains the responsibility of the clinical geneticist or molecular pathologist. No autonomous clinical reporting.

Clinician Review

Every classified variant reviewed by a qualified clinician before entering the patient record. VUS findings explicitly flagged for ongoing reclassification monitoring.

Evidence Traceability

Every classification linked to its evidence base — database references, population frequencies, functional predictions, and literature citations. Full audit trail from raw VCF to final clinical report.

Data Sovereignty

Genomic data processed and stored under the same AES-256 encryption, RBAC, and data residency controls as all AXIVIS clinical data. Country-level genomic data residency configurations available.

4
Sequencing Inputs
WGS, WES, targeted gene panels, and pharmacogenomic profiling — all processed natively from VCF format.
5
Classification Tiers
ACMG/AMP five-tier classification with full evidence documentation for every variant.
5
Reference Databases
ClinVar, gnomAD, COSMIC, OMIM, and PharmGKB cross-referenced for every detected variant.
2
Analysis Modes
Germline (hereditary risk, carrier status) and somatic (tumour profiling, therapeutic targets) — unified pipeline.
Who Genomic Intelligence Serves

Clinical Geneticists

Specialists interpreting germline variants for hereditary risk assessment, carrier screening, and predictive testing with AI-augmented classification and evidence aggregation.

Molecular Pathologists

Pathologists analysing somatic tumour profiles for therapeutic target identification, resistance mutation detection, and molecular subtyping to guide oncology treatment.

Prescribing Clinicians

Physicians using pharmacogenomic profiles to optimise drug selection and dosing — with genotype-specific alerts integrated into the prescribing workflow at the point of care.

Advancing medical intelligence.
At the frontier of medicine.

AXV-09 · Radiology AI

AI-assisted diagnostic imaging interpretation.
Structured. Auditable. Clinician-validated.

CNN-based medical imaging analysis across seven modalities with structured report generation, key finding extraction, clinical impression synthesis, and radiologist worklist intelligence. Every output requires qualified clinician review before clinical use.

EU MDR Class IIa · CNN Pattern Recognition · Human-in-the-Loop Mandatory
Deployment Options

Deploy Radiology AI standalone, or as part of the AXIVIS OS.

Standalone · AXV-09
Radiology AI on its own.
Procure Radiology AI as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Radiology AI as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Radiology AI · Live Product

A reading room, not a reporting tool.

What follows is the shipping Radiology AI product — patient-grade interface, clinician-validated workflow, and every feature you see in the screenshots live in active institutional deployments today. The examples below walk through an 8mm solid pulmonary nodule in a 57-year-old patient across every tab a radiologist touches: interpretation, key findings, impression, structured report, serial comparison, DICOM viewing, pending orders, and departmental analytics.

Allergy · Contrast Prior contrast reaction documented (2021)Anaphylaxis to iodinated contrast. Verify with clinical team before proceeding. → acknowledge
Critical · NOF Robert Flanagan · MRN-910321 · XR · ACC-83871Suspected NOF fracture — requires immediate orthopaedic review. 7m ago · 10:58. → escalated
STAT · Unread Margaret Osei · MRN-821045 · CT · ORD-2026-00841Pending STAT CT chest — haemoptysis. Order placed 07:42, unread. Auto-escalated to department head — Dr. M. Chen notified at 07:57. → escalated
SM
Sarah Mitchell
DOB: 1968-04-12 · Age: 57 · MRN-447821
Order
ORD-2026-08421
Accession
ACC-83921
Study Date
2026-03-30
Contrast
IV+
Images
342 img / 8 series
AI Score
94%
AI Flag
CRITICAL
Interpretation Key Findings Impression Structured Report Comparison Metadata DICOM Viewer
Radiology AI interpretation tab — anatomical regions list, full imaging interpretation, prior study comparison, with critical alerts and allergy banner
01 · Interpretation

Anatomical-region prose built to radiologist reporting cadence.

The interpretation surface opens with a structured list of the anatomical regions covered by the study — Right Lower Lobe, Left Lung, Mediastinum, Lymph Nodes, Liver, Spleen, Kidneys, Adrenals, Pelvis — followed by a full imaging interpretation written in the prose and rhythm of a dictated report.

Prior study comparison is surfaced inline. PRIOR STUDY AVAILABLE · SCAN 1 OF SERIES · CT · 2023-11-14 · ACC-71204. Each prior finding compared against current: No pulmonary nodule → N/A → NORMAL, Centrilobular emphysema → Mild → REVIEW. The radiologist sees delta, not just current state.

Clinical ContextAllergy banner pinned at the top of the report: anaphylaxis to iodinated contrast (2021) — verify with clinical team before proceeding. Acknowledgement logged to the audit trail.
Critical Rail2 CRITICAL ALERTS — ACTION REQUIRED: suspected NOF fracture requiring immediate orthopaedic review; unread STAT CT auto-escalated to department head after 20 minutes unactioned.
Prior StudyIngestion of prior imaging from the same patient — finding-by-finding comparison table with status chips (NORMAL · REVIEW · CRITICAL) rendered directly under the interpretation.
Cortex ContextThe Cortex side panel reads the report and suggests three clinically-grounded follow-up questions — significance of critical findings, potential causes of the nodule, indicated follow-up imaging.
Key Findings overview with WORSENING / STABLE / NORMAL chips, organ tagging, ICD-10 codes, confidence bars, prior-current comparison
02 · Key Findings

Every finding rendered as a clinical record card.

Findings are not extracted as a flat list. Each is rendered as a structured card: clinical title, interval-change chip (WORSENING · STABLE · NORMAL), severity status (CRITICAL · REVIEW · NORMAL), organ tag, ICD-10 code, clinical description, AI confidence bar, and a dense PRIOR → CURRENT comparison row with actual measurements.

The first card on this study: Solid pulmonary nodule · WORSENING · CRITICAL · Right Lower Lobe · ICD-10 R91.1. 91% confidence. Prior: Not identified. Current: 8mm × 6mm. The card writes itself from CNN detection plus prior-study delta — the radiologist sees, confirms, signs.

Status TaxonomyThree-tier severity (CRITICAL · REVIEW · NORMAL) combined with three-tier interval-change (WORSENING · STABLE · IMPROVING) — nine distinct clinical postures rendered as chip pairs.
Structured CodingEvery finding tagged with anatomical organ and ICD-10 code — queryable for cohort search, billing, and epidemiological analysis without free-text extraction.
Confidence LayerPer-finding AI confidence bar (91% · 88% · 97% in this study) — calibrated against ground-truth annotated datasets. Low confidence surfaces for human review first.
Serial ContextTop ribbon declares: Serial study (scan 2) · comparison data available. Click any finding for before/after analysis at the measurement level.
Finding detail drill-down — 99% AI detection confidence, measurements panel for solid organs, descriptors, clinical significance, Ask AI action
03 · Finding Detail

Drill into a finding and see the full measurement ledger.

Clicking any finding card opens a dedicated page for that finding alone. Title, organ tag, severity chip at the top. AI Detection Confidence rendered as a proportional bar — 99% for unremarkable solid organs in this view.

Measurements surface structured, organ-by-organ: Liver 14.5cm craniocaudal — upper limit of normal · Spleen 9.8cm — normal · Kidneys Right 10.2cm, Left 10.6cm — normal · Adrenals Bilateral limbs ≤5mm — normal. Descriptors render below in report-ready prose. A clinical significance block renders the clinical posture in one line. "Ask AI about this finding" opens a Cortex context bound to this specific finding.

Measurement ExtractionOrgan-specific measurement vocabulary (craniocaudal, short-axis, limb thickness) extracted from the CNN pipeline and rendered in the clinical format the radiologist already uses.
Reference RangesEvery measurement tagged against normal reference: upper limit of normal · normal. Deviations surface as non-normal chips and route to the significance summary.
Finding-Bound CortexCortex context scoped to this single finding — not the full report. The assistant answers with the measurements, descriptors, and prior comparison for this finding only.
Back Navigation← All findings breadcrumb returns to the overview grid. Every drill-down is a first-class route — shareable, back-button-safe, audit-logged.
Impression tab with study information, technique, impression citing Fleischner Society 2017, and findings written by anatomical region
04 · Impression

The dictated report, generated before the radiologist reaches for the microphone.

The Impression tab renders the full structured report in the format used across radiology departments worldwide. STUDY INFORMATION — modality, body parts examined, study description, date. TECHNIQUE — reconstruction detail: "Multidetector CT chest, abdomen and pelvis acquired in portal venous phase following 80ml IV iohexol 350 at 3ml/s. Reconstructions at 1.25mm axial, coronal and sagittal."

The IMPRESSION block cites guidelines by name: "Per Fleischner Society 2017 guidelines for solid nodules ≥6mm in a high-risk patient, follow-up LDCT at 3 months is recommended to assess for interval growth." The FINDINGS block reads as structured report text — Lungs and airways, Mediastinum, Abdomen and pelvis — in the exact conventions the attending radiologist dictates in.

Guideline BindImpressions cite the specific guideline that drives them — Fleischner Society 2017, GOLD 2024, BI-RADS 5, TI-RADS, Lugano, ASPECTS — by name and year.
Technique CaptureContrast volume, agent, infusion rate, and reconstruction parameters captured in the technique block — written in the style used for PACS reporting.
Anatomical OrderFindings blocks ordered by anatomical region following radiologist-dictation convention — Lungs and airways first, then Mediastinum, then Abdomen and pelvis.
Sign-Off ReadyEvery block editable inline. Approval with Approve & Sign locks the report, triggers PDF export, and routes signed output to PACS and the referring clinician.
Structured Report with six reporting templates — Free Text, Fleischner (Nodule), Lugano (Lymphoma), ASPECTS (Stroke), BI-RADS (Breast), TI-RADS (Thyroid)
05 · Structured Report

Six specialty templates, each bound to the guideline that governs it.

Reporting templates are not free text with headings. Each is a structured schema bound to the governing international guideline: Free Text, Fleischner (Nodule), Lugano (Lymphoma), ASPECTS (Stroke), BI-RADS (Breast), TI-RADS (Thyroid).

Free Text Fleischner (Nodule) Lugano (Lymphoma) ASPECTS (Stroke) BI-RADS (Breast) TI-RADS (Thyroid)

Clinical history surfaced from the referral: "Known smoker, 35 pack-years. Presenting with haemoptysis ×3 days. Prior CT chest 2023 unremarkable." Findings, Impression, and Recommendations each carry their own Auto-fill AI and Dictate actions. The Recommendations block carries an Auto-Guidelines toggle that pulls from the Fleischner Society 2017 and GOLD Guidelines 2024 directly.

Template SchemaEach template is a structured schema with required fields. Switching template restructures the report — not the content.
Auto-fill AIPer-block AI completion pulls from the imaging analysis and the radiologist's prior dictation style. Dictate action opens a voice-to-text lane for any block.
Auto-GuidelinesRecommendations block auto-populates with the clinically relevant guideline text — "Fleischner Society 2017: Follow-up LDCT at 3 months", "GOLD Guidelines 2024: Pulmonary function testing".
Finalise & SignSave Draft preserves work without locking. Finalise & Sign Report applies the clinician's digital signature, triggers PDF export, and dispatches the report.
Serial scan comparison with prior and current studies side by side, interval change highlighted, AI predictive analysis
06 · Comparison

Synchronised side-by-side with interval change analysis.

The Comparison tab renders the prior and current studies in parallel viewports with synchronised navigation. PRIOR STUDY · CT · 2023-11-14 · ACC-71204 on the left. CURRENT STUDY · CT · 2026-03-30 · ACC-83921 on the right. Window and level values (W:350 · L:50) surfaced in the viewport chrome.

The new 8mm nodule is annotated directly on the current study — "8mm NEW" label pointing to a dashed-red ROI in the right lower lobe. Below, INTERVAL CHANGE · FINDING BY FINDING renders every finding with a prior/current delta. In this study: "Solid pulmonary nodule · Right Lower Lobe · WORSENING · CT Chest Nov 2023: Not identified → Current: 8mm × 6mm · NEW — interval development over 28 months". An AI PREDICTIVE ANALYSIS block follows.

Scroll-SyncPrior and current navigate in lockstep. Scroll one, scroll both — at the same anatomical level, at the same window/level.
Interval ChangeEvery finding carries a interval-change status. NEW, STABLE, IMPROVED, WORSENING — with measurement deltas rendered in the prior/current cells.
ROI AnnotationAI-detected new findings annotated on the current study with dashed-red ROI and size label — visible against the prior with no mark.
Predictive LayerAI predictive analysis estimates interval-development window — "NEW — interval development over 28 months" — and links to the governing guideline for follow-up cadence.
DICOM viewer with full tool palette and AI overlay showing ROI circles with 91% and 88% confidence on CT chest
07 · DICOM Viewer

The AXIVIS Imaging Suite, embedded in the report.

From any report tab, DICOM Viewer opens the full AXIVIS Imaging Suite on the underlying study — dual-canvas rendering, full measurement palette, window/level presets, multi-planar reconstruction.

In this view, the CT chest is rendered with the AI overlay engaged (AI ON). Two AI detection markers visible directly on the image: AI 91% · Right Lower Lobe and AI 88% · Bilateral Lungs — rendered as semi-transparent coloured ROIs anchored to the CNN detection coordinates. The radiologist can toggle the AI overlay off at any time, adjust confidence threshold, and interrogate any region with the ROI AI tool.

Tool PaletteW/L · Pan · Zoom · Ruler · Angle · Ellipse · Rect · Free · Arrow · Text · Cobb · Bidir · Cal · Mag · ROI AI · Flip H · Flip V · Rotate — the full measurement primitives used in the Imaging Suite.
W/L PresetsAuto · Lung · Soft · Bone · Brain · Liver · Mediastinum — modality-appropriate window/level presets one click away.
AI OverlayAI detections rendered as coloured ROI overlays on the image with confidence scores. Toggle on/off, adjust threshold, or interrogate a custom region with ROI AI.
PACS LoopLoad · PACS · Prior — load local DICOM, query PACS via QIDO-RS, or pull prior studies scroll-synced with the current series.
Radiology Performance Dashboard — studies today, average turnaround, AI accuracy, AI flags, critical alerts, hourly throughput, modality turnaround, radiologist workload
08 · Analytics

A departmental dashboard built for governance, not vanity metrics.

Five departmental KPIs render at the top of the analytics surface: Studies Today, Avg Turnaround with per-target comparison, AI Accuracy with radiologist override rate, AI Flags with critical subset, and Critical Alerts unacknowledged count.

Studies Today
41
Currently reporting
Avg Turnaround
5.6m
Target 8m
AI Accuracy
96.2%
Override rate: 4.9%
AI Flags
8
3 critical
Critical Alerts
3
Unacknowledged

Hourly Throughput stacked by modality — CT, MRI, XR, PET segments from 06:00 onward. Turnaround Time by Modality rendered as horizontal bars with per-modality target: CT 4.1m · MRI 6.8m · XR 2.4m · PET 9.2m · Department Average 5.6m.

CT
4.1m
Target 8m
MRI
6.8m
Target 8m
XR
2.4m
Target 8m
PET
9.2m
Target 12m
Radiologist WorkloadPer-radiologist table: online status, studies read, critical flags, average TAT, performance tier (Excellent · On target). Dr. Harwood · 14 studies · 4.2m TAT · Excellent.
STAT EscalationSTAT Escalation Monitor runs at the bottom of the dashboard — auto-assigns after 20 min unactioned, with per-STAT escalation path visible to the on-call lead.
Override TrackingAI override rate computed from every declined or refined recommendation. Override spikes trigger a retraining review flag for the governance loop.
Time WindowToggle TODAY · WEEK · MONTH at the top right. All KPIs, charts, and tables recompute against the selected window.
Pending Orders queue with Total/Pending/In Progress/Scheduled/Completed/Rejected counters, filterable order table with modality chips and action buttons
09 · Pending Orders

The full radiology order queue in one department-level view.

Six order-state counters render at the top: Total Orders 8 · Pending 4 · In Progress 1 · Scheduled 1 · Completed Today 0 · Rejected 1. Search by patient, order number, or department. Filter by modality, priority, or status.

Each row carries the full institutional order record: Order No (RO-2026-00812), placed timestamp, priority chip (URGENT · ROUTINE · STAT), patient identity, modality chip (MAMMO · PET · NM · MRI · US · CT), protocol, clinical indication, assigned radiologist, target date, and per-row actions. Overdue orders carry an OVERDUE flag rendered in red below the status chip.

Order TaxonomyModality-bound order types — MAMMO bilateral + tomosynthesis, PET whole body FDG, NM bone scan (SPECT-CT), MRI lumbar spine, US abdomen + pelvis, CT chest with contrast — each with its own scheduling rules.
Clinical IndicationEvery order carries a full clinical indication — "DLBCL post cycle 6 R-CHOP. Restaging PET for treatment...", "Known prostate carcinoma (Gleason 8). Elevated PSA...", "Haemoptysis × 2 weeks. Ex-smoker. ?Lung carcinoma." — written by the referring clinician.
Row ActionsTake Order · Schedule · Assign · Upload · Reject — five institutional actions per pending row. Upload Study · Complete for scheduled rows.
Department RoutingEach order labelled with originating department — Breast Surgery, Haematology, Oncology, Orthopaedics, Gynaecology, Respiratory Medicine — for analytics segmentation and workload distribution.
Supported Modalities

Seven imaging modalities. One intelligence layer.

Each modality processed by dedicated CNN models trained on annotated clinical imaging datasets.

CT
Computed Tomography
Multi-slice volumetric analysis with anatomical segmentation
MR
MRI
Multi-sequence tissue characterisation and lesion detection
XR
X-Ray
Chest, skeletal, and abdominal plain film interpretation
US
Ultrasound
Real-time sonographic analysis with measurement extraction
PT
PET/CT
Metabolic activity quantification with anatomical correlation
MG
Mammography
Screening and diagnostic with BI-RADS classification support
DX
DEXA
Bone mineral density assessment and fracture risk analysis
Imaging Analysis Pipeline

Pattern recognition trained on clinical imaging reality.

AXIVIS Radiology AI processes uploaded imaging studies through modality-specific convolutional neural network models. The system identifies anatomical structures, detects potential abnormalities, characterises findings by location and morphology, and generates structured interpretations with confidence scoring. Every AI finding is annotated directly on the imaging study for immediate radiologist review.

Real-Time AI Annotation

Confidence-Scored Finding Markers

Each detected finding is overlaid on the imaging study with a confidence percentage. Radiologists see AI markers (AI 88%, AI 91%) directly on the scan, enabling rapid visual correlation between AI detection and anatomical context.

Multi-Window Viewing

Lung · Soft · Bone · Brain · Liver · Mediastinum

One-click window presets optimised for each tissue type. Custom window width and level adjustment with real-time slider controls.

Integrated Reporting

Interpretation · Key Findings · Comparison · Structured Report

Four-tab report generation accessible directly from the viewer toolbar. AI-generated structured reports with clinician sign-off before finalisation.

AXIVIS DICOM Viewer · AI Analysis Active
Live
AXIVIS DICOM Viewer with AI Analysis
Structured Radiology Report
AI-Generated · Pending Review
Interpretation
Key Findings
Impression
Metadata
Interpretation
CT Chest with Contrast — Axial, Coronal, Sagittal Reconstructions. Lungs: A 14mm spiculated nodule is identified in the right lower lobe (segment 6), new compared to the prior study dated 2025-01-15. No cavitation. No surrounding ground-glass halo. A 4mm ground-glass opacity in the left upper lobe is unchanged. Mediastinum: A subcarinal lymph node measures 12mm in short axis, borderline by size criteria. No pathological hilar lymphadenopathy. Pleura: No effusion. No pneumothorax.
Clinical Impression
New spiculated RLL nodule requiring further characterisation. Recommend PET/CT for metabolic assessment. If PET-avid, CT-guided biopsy for tissue diagnosis. Subcarinal lymph node warrants follow-up at next imaging interval. Stable LUL ground-glass opacity — continue surveillance per Fleischner criteria.
Structured Reporting

Complete radiology reports generated. Clinician edits before finalisation.

AXIVIS generates structured radiology reports across four output tabs: Interpretation (systematic anatomical review with findings), Key Findings (extracted list with location, measurements, and classification), Impression (clinical summary with recommendations), and Metadata (study parameters, comparison studies, technical quality). The reporting clinician reviews, edits, and approves the generated report before it enters the patient record. No report is finalised without clinician sign-off.

Four-Tab Report Structure

Interpretation · Key Findings · Impression · Metadata

Systematic separation of descriptive findings, extracted data points, clinical synthesis, and technical parameters.

Full Editing Capability

Clinician Retains Complete Control

Every section of the generated report is editable. The radiologist or reporting physician has full authority to modify, correct, or override any AI-generated content.

A CT chest is uploaded. AXIVIS identifies three findings across lungs and mediastinum, generates a structured interpretation, extracts key findings with measurements and classifications, and drafts a clinical impression with follow-up recommendations. The radiologist reviews the report, adjusts the impression wording, and signs off. Total time from upload to signed report: under 3 minutes.
Interactive DICOM Viewer · AXV-16

Browser-based medical imaging. No external software required.

AXIVIS includes a fully integrated DICOM viewer within the web platform. Clinicians open and review imaging studies directly in the browser with professional-grade visualisation tools. ROI AI analysis enables region-specific AI interpretation with confidence-scored annotations. The toolbar provides ruler measurement, angle calculation, pan, zoom, window/level adjustment, and invert capabilities — all within a single unified interface.

ROI AI Analysis

Region of Interest · AI-Powered Interpretation

Select any region on the imaging study for targeted AI analysis. The system returns confidence-scored findings with anatomical localisation and clinical significance assessment.

Professional Toolbar

1fr · W/L · Pan · Zoom · Ruler · Angle · HU · ROI AI

Complete diagnostic imaging toolkit accessible from a single toolbar row. Snapshot capture and study loading for multi-series navigation.

Clinical Workflow Integration

Worklist · STAT · Peer Review · MDT · DICOM · PDF Report · Approve & Sign

Full clinical workflow from worklist triage through peer review, multidisciplinary team submission, and final approval with digital signature.

AXIVIS DICOM Viewer · ROI AI Enabled
STAT
AXIVIS DICOM Viewer with ROI AI
AI Triage Worklist
STAT
CT Head — Acute stroke query
62F · ED · Onset 45 min ago
0.94
URGENT
CT Chest — New pulmonary nodule
58M · Outpatient · Prior comparison available
0.81
URGENT
MRI Brain — Suspected metastasis
71M · Oncology · Known primary lung
0.77
ROUTINE
X-Ray Chest — Pre-op assessment
44F · Surgical · No prior imaging
0.23
ROUTINE
DEXA — Osteoporosis screening
67F · Rheumatology · Baseline study
0.12
AI Triage Intelligence

Priority scoring. STAT flagging. Worklist optimisation.

Radiology departments process hundreds of imaging studies daily. AXIVIS assigns priority scores based on initial AI assessment of clinical urgency — STAT cases flagged immediately, urgent cases elevated, routine cases queued appropriately. The radiologist's worklist is optimised so the most clinically critical studies are reviewed first. Priority scoring is advisory — the radiologist retains full authority to re-prioritise at any time.

Acuity-Based Scoring

0.0 — 1.0 Priority Scale

Every study receives a numerical priority score based on clinical context, referral urgency, AI preliminary assessment, and patient acuity indicators.

STAT Detection

Automatic Escalation for Critical Findings

Studies with findings suggestive of acute pathology (stroke, PE, pneumothorax, fracture) flagged for immediate radiologist attention. Real-time push notifications to mobile app.

Department Dashboard

Workload Visibility

Department heads see real-time study volumes, completion rates, turnaround times, and pending worklist depth across all radiologists.

Longitudinal Intelligence

Lesion tracking across imaging studies. Progression quantified over time.

Incidental findings and known lesions require monitoring across sequential imaging studies. AXIVIS automatically compares measurements from current and prior studies, calculates volumetric change, and applies monitoring criteria (Fleischner, Lung-RADS, BI-RADS) to generate structured follow-up recommendations. The system tracks the trajectory — growth rate, doubling time, stability — not just the latest measurement.

Automated Comparison

Prior Study Registration

Current findings automatically matched against prior imaging studies for the same patient. Measurement delta calculated with percentage change and absolute difference.

Monitoring Criteria

Fleischner · Lung-RADS · BI-RADS

Evidence-based monitoring guidelines applied automatically. Follow-up interval recommendations generated based on finding characteristics and change trajectory.

Growth Kinetics

Doubling Time · Volume Change · Trajectory

Quantitative growth analysis for serial lesion measurements. Exponential growth patterns flagged for accelerated clinical review.

Lesion Tracking
Longitudinal
Jan 2025
8mm
Apr 2025
11mm
Jul 2025
14mm
Growth rate: +75% over 6 months · Doubling time: ~8 months · Exceeds Fleischner threshold
Clinical Safeguards

Every imaging output advisory. Every finding clinician-validated.

Advisory Output

All AI-generated findings and reports are decision support. Final radiology interpretation remains the responsibility of qualified radiologists and physicians. No autonomous diagnosis.

Clinician Editing

Every section of the generated report is fully editable. The reporting clinician reviews, corrects, and approves before any report is finalised or enters the patient record.

Model Traceability

Every AI inference logged with the specific model version, input study reference, output generated, confidence scores, and the reviewing clinician who signed off.

Data Integrity

Original DICOM imaging data preserved without modification throughout the analysis, viewing, and reporting workflow. Annotations are overlaid, never embedded.

7
Imaging Modalities
CT, MRI, X-ray, Ultrasound, PET/CT, Mammography, DEXA — all processed natively.
33M+
Training Image Pairs
Vision intelligence pre-trained across radiology, histopathology, ophthalmology, and dermatology datasets.
<3min
Upload to Signed Report
From imaging upload through AI analysis, structured report generation, clinician review, to final sign-off.
4
Report Output Tabs
Interpretation, Key Findings, Impression, and Metadata — structured for clinical workflow.
Who Radiology AI Serves

Radiologists

Primary reporting clinicians using AI-assisted analysis for structured interpretation, finding extraction, and report generation across all supported imaging modalities.

Referring Physicians

Clinicians ordering imaging studies who access structured reports, key findings, and clinical impressions within the patient record and through AXIVIS Cortex.

Radiology Departments

Department leadership with visibility into study volumes, turnaround times, triage effectiveness, and workflow compliance across the entire imaging operation.

Advancing medical intelligence.
At the frontier of medicine.

AXIVIS Platform · Module 16

Imaging Suite.
A hospital-grade diagnostic workspace.

Zero-footprint, browser-delivered DICOM workstation with embedded clinical reasoning. Dual-canvas rendering, twelve measurement primitives, multi-planar reconstruction, and AXIVIS Vision per-frame analysis — integrated with PACS through standard DICOMweb.

ClassificationDiagnostic Workspace
DeploymentZero-footprint
Mount/doctor/imaging
Version2026.04 · Rev A
Live · AI Analysis AXIVIS Imaging Suite — CT head with ROI AI region analysis panel
Deployment Options

Deploy AXIVIS Imaging Suite standalone, or as part of the AXIVIS OS.

Standalone · AXV-16
AXIVIS Imaging Suite on its own.
Procure AXIVIS Imaging Suite as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy AXIVIS Imaging Suite as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
At a Glance

Engineered for the institutional reading room.

Modalities
8+ supported
CT · MRI · X-Ray · Ultrasound · PET · Nuclear Medicine · Mammography · generic DICOM — each with modality-specific hanging protocols and window/level presets.
Measurement Tools
12 primitives
Length, Angle, Ellipse, Rectangle ROI, Freehand, Arrow, Text, Calibration, Cobb angle, Bidirectional RECIST, Segmentation brush, and ROI-to-AI region interrogation.
Endpoint Footprint
0 MB
No endpoint software. No driver dependencies. Delivered through a modern browser with WebGL2 — compatible with standard hospital workstations.
DICOM Exports
4 formats
Structured Report (SR), Secondary Capture (SC), Key Object (KO), and Grayscale Softcopy Presentation State (GSPS) — all round-tripping to PACS.
Frame Cache
200 frames
In-browser LRU with memory-bounded budget, abortable prefetch controller for adjacent frames, and auto-subsampled preview tier for large frames.
Viewport Layouts
9 layouts
1×1, 1×2, 1×3, 2×2, 2×3, 3×3 with click-to-activate selection. Prior-study comparison uses a second Frame Vault with scroll-sync.
Architecture · Five Pillars

One workspace, five clinical primitives.

The Imaging Suite is not a single component. It is five tightly-integrated primitives, each engineered for its clinical purpose and deployed as a single workflow.

01
Viewer
Hospital-grade DICOM engine
Dual-canvas rendering (CPU + optional WebGL2 GPU) with HiDPI awareness, dirty-hash render loop, full W/L control, multi-planar reconstruction, and configurable slab thickness with MIP, MinIP, and Average projection modes.
02
Tools
Twelve measurement primitives
Length, angle, ellipse area, rectangular ROI, freehand, arrow, text, calibration, Cobb angle (orthopedic), bidirectional RECIST (oncology), segmentation brush, and ROI-to-AI region interrogation. Full undo/redo.
03
Cortex
Embedded clinical reasoning
Streaming synthesis with intent classification and safety-first response hierarchy. Patient demographics, prior findings, frame-level observations, and supporting documents auto-injected — reads like an attending consult.
04
Vision
Per-frame AI analysis
Frame-level analysis on demand or automatically as the radiologist scrolls. Returns severity classification, structured findings, summary, and confidence. Results persist in study-level frame memory.
05
Vault
Persistent study storage
Studies, presentations, measurements, and key images persist to encrypted institutional storage with presigned access. Annotations round-trip through DICOM GSPS. Audit-friendly study lifecycle with soft-delete.
In Practice · Reading Room

The interface a radiologist actually uses.

The screenshots below are from a live institutional deployment. Patient identifiers redacted; interface unchanged.

Reading workspace — W/L controls, measurement toolbar, frame navigation
Reading Workspace · CT Head
Full tool palette surfaced above the canvas: W/L · Ruler · Angle · Ellipse · Rectangle · Freehand · Arrow · Cobb · Bidirectional · Calibrate · Magnify · ROI AI. Bottom bar shows Cortex online status.
Vision analysis — NORMAL classification with detailed findings
AXIVIS Vision · Frame Analysis
Per-frame AI returns severity (NORMAL · NOTABLE · CRITICAL), structured findings across calvarium, brain parenchyma, ventricles, and soft tissues, plus confidence scoring. Persisted to study frame memory.
Rendering Pipeline

Sub-frame efficiency through a dirty-hash render scheduler.

Every frame flows through a six-stage pipeline. Redundant draws are elided. State changes are the only trigger for a render pass.

01 · INGEST
DICOM Parse
In-browser Part 10 parser · off-thread worker decode · priority queuing
02 · CACHE
Frame Vault
LRU 200 frames · memory-bounded budget · abortable prefetch
03 · TRANSFORM
W/L · Pan · Zoom
Mutable refs · bicubic interp >2× zoom · flip/rotate
04 · RASTER
Canvas 2D · WebGL2
R32F texture · capability probe · graceful fallback
05 · OVERLAY
Annotation Canvas
Dedicated overlay plane · 12 primitives · undo/redo stack
06 · COMMIT
HiDPI Display
devicePixelRatio scaled · dirty-hash elides no-ops
Supported Modalities

Every modality ships with appropriate hanging protocols.

Default window/level, modality-specific layout intelligence, and colour-coded UI for instant visual identification on multi-study worklists.

CT
Computed Tomography
Chest · Abdomen · Brain
W:400 · L:40
2×2 layout — Lung · Mediastinum · Bone · Detail
MRI
Magnetic Resonance
T1 · T2 · FLAIR · DWI
W:400 · L:200
Configurable per series
XR
X-Ray · Radiograph
Projection imaging
W:2000 · L:200
Single viewport
US
Ultrasound
B-mode · Doppler
W:400 · L:200
Single viewport
PET
Positron Emission
FDG · PSMA · Amyloid
W:800 · L:400
Single + fusion-ready
NM
Nuclear Medicine
Bone scan · SPECT
W:600 · L:300
Single viewport
MG
Mammography
CC · MLO bilateral
W:4000 · L:2000
Auto bilateral 1×2
DCM
Generic DICOM
Any DICOM Part 10
Auto W/L detection
Configurable layout
Worklist Intelligence

A worklist that tells you where to look first.

Hero statistics at the top: Studies, Critical, Notable, Frames Analyzed, Normal. Modality filters, severity sort, and bulk operations. 20 studies per page with status chips — URGENT, STAT, AI REVIEWED, PENDING REVIEW.

AXIVIS Imaging Studio — worklist dashboard with recent studies
Interoperability

PACS-native by standard protocols.

Connects to any standards-conformant DICOMweb endpoint. No middleware. No proprietary driver. Full bidirectional DICOM GSPS for presentation state round-trip.

Inbound · Read
DICOM Part 10 Single-frame and multi-frame studies · full metadata
Inbound · Query
DICOMweb QIDO-RS Native study and series query from integrated browser
Inbound · Retrieve
DICOMweb WADO-RS Frame retrieval through DICOMweb browser
Inbound · Import
GSPS Import Load grayscale softcopy presentation state from external PACS
AXIVIS
Imaging Suite
Gateway
Outbound · Export
DICOM SR Structured Report with measurements and key images
Outbound · Export
DICOM SC Secondary Capture from viewport state
Outbound · Export
DICOM KO · GSPS Key Object notes and grayscale presentation state
Outbound · Upload
DICOMweb STOW-RS Server-side proxied upload to PACS
Measurement Palette

Twelve primitives engineered for diagnostic precision.

DEFAULT DRAG
Length
Distance in mm · uses DICOM pixel spacing
3 POINT
Angle
Three-point angle measurement in degrees
Shift+E
Ellipse Area
Elliptical region of interest with area
DRAG
Rectangle ROI
Rectangular region for bounded measurement
Shift+F
Freehand
Freehand region outline for irregular morphology
Shift+A
Arrow
Directional annotation arrow
Shift+C
Cobb Angle
Four-point Cobb angle for orthopedic spinal assessment
Shift+B
Bidirectional
RECIST-style bidirectional diameter for oncology
Shift+L
Calibrate
Pixel spacing calibration for non-standard studies
Shift+T
Text
Text annotation with clinician attribution
BRUSH
Segmentation
Brush, eraser, and threshold with opacity / label management
ROI → AI
ROI Interrogation
Region selection routed to AXIVIS Vision for bounded interpretation
Keyboard

For the radiologist who never lifts their hands.

AToggle AI sidebar
SOpen study synthesis
CCortex · imaging context
PToggle panel
EscReturn to dashboard
WheelScroll frames
Ctrl+WheelZoom in / out
MiddlePan image
Public API Surface

Designed for institutional integration.

All endpoints mounted under the gateway prefix. JWT-authenticated. Full OpenAPI specification available on request.

Core AI

POST/axivis-ai/analyze
POST/axivis-ai/analyze-frame
POST/axivis-ai/triage
POST/axivis-ai/endoscopy-analyze
POST/axivis-ai/cortex · SSE

Studio & Annotations

POST/axivis-ai/studio/upload
GET/axivis-ai/studio/studies
GET/axivis-ai/studio/study/{id}/files
POST/axivis-ai/studio/study/{id}/memory
POST/axivis-ai/imaging-annotations/save
GET/axivis-ai/imaging-annotations/load/{uid}

DICOM Export

POST/axivis-ai/export/dicom-sr
POST/axivis-ai/export/dicom-sc
POST/axivis-ai/export/dicom-kin
POST/axivis-ai/export/dicom-pr
POST/axivis-ai/import/gsps
POST/axivis-ai/pacs/stow
Safety & Governance Posture

An augmentation layer for credentialed clinicians.

The Imaging Suite is designed explicitly as an augmentation layer — not an autonomous diagnostic device. Every imaging AI output is framed as a preliminary reading requiring clinical correlation.

Clinical Positioning
Augmentation layer for credentialed clinicians. Not an autonomous diagnostic device. All AI output explicitly presented as preliminary, requiring clinician validation.
Safety Surfacing
Drug interactions, contraindications, and urgent findings flagged at the top of every Cortex response. Most dangerous differential first; most urgent action first.
Authentication
Hybrid JWT-based clinician sessions with secure session fallback. HMAC verification for all background callbacks — workers cannot be impersonated.
Output Sanitisation
Three-layer scrubbing — server, gateway, and client. No internal model identifiers ever reach the clinical UI. Markdown stripping for clean clinical prose.
Data Lifecycle
Soft-delete model preserves study history with full audit trail. Presigned vault access. Configurable data residency region for jurisdictional compliance.
Study Integrity
Local fallback storage activates automatically on transient vault failure. Uploads are never lost — replay to durable vault occurs on recovery.

Read a study. Then talk to us.

The Imaging Suite is deployed in clinical environments today. Schedule a procurement-grade walkthrough with the AXIVIS team — DICOM workflow, PACS integration, security posture, and institutional licensing covered end-to-end.

AXV-05 · Patients

The complete patient record.
Unified. Intelligent. Actionable.

Centralised patient records with AI-generated health insights, diagnostic ordering suggestions, and medication safety checking. Twelve integrated tabs spanning clinical overview, diagnostic reports, imaging, genomics, oncology, prescriptions, and the full patient timeline. Every module's output converges in the patient record.

EU MDR Class IIa · 12 Clinical Tabs · Cross-Module Integration · AI Health Insights
Deployment Options

Deploy Patients Portal standalone, or as part of the AXIVIS OS.

Standalone · AXV-12
Patients Portal on its own.
Procure Patients Portal as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Patients Portal as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Unified Patient Interface

Twelve tabs. One patient. Every data domain.

The patient record is the convergence point for every AXIVIS module. Imaging, genomics, laboratory, oncology, prescriptions — all accessible from a single tabbed interface.

Overview
Reports
Appointments
Orders
Medical History
Billing
Notes
Oncology
Genetics
AI Assistant
Prescriptions
Logs
SM

Sarah Mitchell

DOB: 1968-04-12
Age: 57
MRN: 447821
NHS: 943 221 8847
GP: Dr. A. Patel
PENICILLIN ALLERGY
TYPE 2 DIABETES
BRCA1 POSITIVE
AI Health Insights

The patient record that reasons. Not just records.

Traditional patient records are passive repositories — they store data without analysing it. AXIVIS applies AI across the complete patient dataset to generate health insights: risk patterns identified from longitudinal data, medication interactions flagged from the current prescription list, diagnostic suggestions based on presenting symptoms cross-referenced with medical history, and screening reminders triggered by age, sex, and risk profile. Every insight is advisory and requires clinician review before action.

Cross-Data Pattern Recognition

Labs + Imaging + Genomics + History

AI analyses across data types simultaneously — rising HbA1c from labs, declining eGFR from renal function, BRCA1 status from genomics, and family history from the medical record. Patterns that span data silos are surfaced automatically.

Proactive Risk Identification

Before Symptoms → Before Complications

Risk signals identified from data trajectories before clinical symptoms manifest. Deteriorating renal function flagged for nephrology review. Rising inflammatory markers correlated with imaging findings. Pharmacogenomic interactions detected before prescribing.

Screening Intelligence

Age · Sex · Risk Profile · Family History

Automated screening recommendations based on national guidelines, adjusted for individual risk profile. BRCA1 carriers receive enhanced breast screening reminders. Diabetic patients receive retinal and renal screening prompts.

AI Health Insights · Sarah Mitchell
6 insights · 2 urgent
Renal Function Declining
eGFR has declined from 72 to 48 mL/min over 12 months. Creatinine rising. Current trajectory suggests Stage 3b CKD within 6 months. Nephrotoxic medications on current prescription list (Metformin — requires eGFR review).
URGENT — Nephrology referral + medication review
BRCA1 Carrier — Enhanced Screening Due
BRCA1 c.5266dupC pathogenic variant confirmed (AXV-12). Last breast MRI: 14 months ago. NICE guidelines recommend annual breast MRI for BRCA1 carriers aged 30-69. Screening interval exceeded.
URGENT — Schedule breast MRI + risk-reducing surgery discussion
Glycaemic Control Deteriorating
HbA1c rising: 6.6% → 7.2% over 12 months. Current medications: Metformin 1g BD, Gliclazide 80mg OD. Above NICE target of 6.5%. Consider medication intensification or specialist diabetes review.
REVIEW — Diabetes medication adjustment
Inflammatory Marker Elevated
CRP 28 mg/L (reference <5). WBC 14.2 ×10⁹/L. No documented acute illness. Consider occult infection, autoimmune process, or malignancy screening given BRCA1 status and age.
REVIEW — Investigate inflammatory source
Pharmacogenomic Alert — CYP2C19
Patient is CYP2C19 poor metaboliser (*2/*2). Current prescription includes Clopidogrel — reduced efficacy expected. Alternative antiplatelet agent may be required if cardiovascular indication present.
NOTE — Flag for prescribing clinician
Diabetic Retinal Screening Due
Type 2 Diabetes diagnosed 2019. Last retinal screening: 13 months ago. Annual screening recommended per NICE NG28. No documented retinopathy on last assessment.
NOTE — Schedule retinal screening appointment
Module Data Integration Map
AXV-01/02
Clinical search + AI reasoning
AXV-04
Consultation notes + prescriptions
AXV-07
Diagnostic orders + results
AXV-08
Lab results + biomarker trends
AXV-09
Radiology reports + imaging
AXV-10
Procedural reports + video
AXV-11
Oncology pathway + staging
AXV-12
Genomic variants + pharmacogenomics
All module outputs converge in the patient record → AI reasons across the complete dataset
Cross-Module Convergence

Every module feeds the patient record. The patient record feeds every insight.

The patient record is not a standalone module — it is the convergence point for every other module in the platform. Consultation notes from AXV-04 appear in the Notes tab. Diagnostic orders from AXV-07 appear in the Orders tab. Lab results from AXV-08 appear in the Reports tab. Radiology findings from AXV-09 appear in the Reports tab. Oncology pathway data from AXV-11 appears in the Oncology tab. Genomic results from AXV-12 appear in the Genetics tab. The AI reasons across all of this data simultaneously to generate the health insights that appear in the Overview tab.

Twelve Integrated Tabs

Every Clinical Domain Accessible

Overview (AI insights + summary), Reports (lab + imaging + pathology), Appointments (past + upcoming), Orders (diagnostic lifecycle), Medical History (conditions + surgeries + family), Billing, Notes (consultation + clinical), Oncology, Genetics, AI Assistant (conversational), Prescriptions, and Logs (audit trail).

AI Assistant Tab

Patient-Contextual AXIVIS Cortex

The AI Assistant tab provides a conversational interface with AXIVIS Cortex, pre-loaded with the complete patient context. Clinicians ask questions about the patient — "What is the renal function trend?", "Are there any drug-gene interactions?" — and receive responses informed by the full patient dataset.

Medication Safety Intelligence

Every prescription checked. Against the full patient profile.

The Prescriptions tab integrates with AXV-04 Consultation Suite and AXV-12 Genomic Intelligence to provide comprehensive medication safety. Drug-drug interactions checked across the complete medication list. Drug-allergy cross-referencing against documented allergies. Drug-gene interactions flagged from pharmacogenomic profiling. Age, weight, and pregnancy-adjusted dosing verification. Duplicate therapy detection. Renal and hepatic dose adjustments recommended based on current lab results. Every safety check documented and auditable.

Multi-Layer Safety Checking

Interactions · Allergies · Genomics · Dosing · Duplicates

Five simultaneous safety checks on every prescription: drug-drug interactions, allergy cross-reference, pharmacogenomic interactions, dose appropriateness (age/weight/organ function), and duplicate therapy detection.

Renal/Hepatic Dose Adjustment

Lab-Informed Dosing Recommendations

Current eGFR and liver function tests automatically inform dosing recommendations. Medications requiring renal dose adjustment are flagged when eGFR drops below specified thresholds. Metformin flagged at eGFR <30, gentamicin dose calculated from current creatinine clearance.

Current Medications · Safety Check
Metformin 1g BD
ALERT
eGFR 48 — approaching contraindication threshold (eGFR <30). Monitor renal function. Consider dose reduction.
Clopidogrel 75mg OD
PGx
CYP2C19 poor metaboliser — reduced active metabolite. Consider alternative antiplatelet (Ticagrelor/Prasugrel).
Gliclazide 80mg OD
OK
No interactions. No allergy conflict. Standard dosing appropriate.
Atorvastatin 40mg ON
OK
No interactions. SLCO1B1 normal function — standard myopathy risk.
Clinical Safeguards

Every insight advisory. Every action clinician-initiated.

Advisory Insights

All AI-generated health insights are decision support. Clinical action requires qualified clinician review. No automated clinical decisions are made from AI insights.

Data Completeness

AI insights explicitly flag when data is incomplete or when a finding is based on limited information. Confidence indicators reflect data quality and completeness.

Medication Safety

Prescription safety checks are enforced at the point of prescribing. Critical interactions cannot be overridden without documented clinical justification.

Access Governance

Patient record access governed by RBAC with document-level permissions. Every access logged with clinician identity, timestamp, and pages viewed.

12
Clinical Tabs
Overview, Reports, Appointments, Orders, Medical History, Billing, Notes, Oncology, Genetics, AI Assistant, Prescriptions, Logs.
8
Module Integrations
Data from Medical Intelligence, Consultation Suite, Orders, Labs, Radiology, Procedures, Oncology, and Genomics — all in one record.
5
Safety Checks
Drug-drug interactions, allergy cross-reference, pharmacogenomics, dose verification, and duplicate therapy — on every prescription.
Real-time
Insight Generation
AI health insights generated in real time from the complete patient dataset as new data enters the record.
Who The Patient Record Serves

Clinicians

Physicians, specialists, and allied health professionals accessing the unified patient record with AI-augmented insights, cross-module data, and medication safety intelligence at the point of care.

Patients

Patients accessing their own health record through the patient portal — viewing AI health summaries, upcoming appointments, medication lists, and shared clinical documents.

Care Coordinators

Multidisciplinary team members coordinating complex care pathways with full visibility into the patient's cross-departmental clinical activity, diagnostic results, and treatment plans.

Advancing medical intelligence.
At the frontier of medicine.

AXV-03 · Scheduling & AI Scribe

Appointment intelligence.
With AI documentation built in.

Appointment management with integrated AI Scribe. NLP converts consultation dialogue into structured clinical notes in real time — SOAP format, discharge summaries, referral letters. The clinician reviews and approves before any note is finalised. Every consultation documented in under 60 seconds. Every word clinician-validated.

EU MDR Class IIa · NLP Clinical Transcription · SOAP Generation · Clinician Approval Mandatory
Deployment Options

Deploy Scheduling & AI Scribe standalone, or as part of the AXIVIS OS.

Standalone · AXV-11
Scheduling & AI Scribe on its own.
Procure Scheduling & AI Scribe as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Scheduling & AI Scribe as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Appointment Management

Structured scheduling across every department and clinician.

AXIVIS scheduling integrates appointment management directly with the clinical workflow. Appointments link to patient records, clinical history, and prior consultation notes before the patient arrives. Clinicians see their daily schedule with patient context pre-loaded. Reception staff manage bookings through a dedicated dashboard. Urgent and walk-in appointments slot into the schedule with priority flags. Every appointment becomes the entry point for the AI Scribe — when the consultation begins, documentation starts automatically.

Pre-Consultation Context

Patient History · Prior Notes · Outstanding Results

Before the appointment starts, the clinician sees the patient's relevant history, prior consultation notes, outstanding test results, and active medication list. No chart review required — context is pre-loaded.

Multi-Department Views

GP · Specialist · Radiology · Endoscopy · Surgical

Department-specific scheduling views with configurable slot durations, appointment types, and capacity limits. Cross-department booking for referrals and multi-disciplinary appointments.

Patient Notifications

Booking · Reminder · Arrival · Follow-Up

Automated patient communications at booking confirmation, 48-hour reminder, day-of-appointment arrival instructions, and post-consultation follow-up with any actions or prescriptions.

Dr. K. Okonkwo · Clinic Schedule
◀Wednesday 8 April▶
MON
TUE
WED
THU
FRI
08:00
S. Mitchell
A. Patel
M. Rahman !
J. Williams
Available
09:00
L. Chen
Available
H. Sayed
D. O'Brien
F. Ahmed
10:00
R. Kovac
T. Nguyen
B. Taylor
Available
Walk-in !
11:00
Available
K. Santos
P. Volkov
E. Garcia
C. Park
AI Scribe · Live Transcription
DR
Good morning Mrs Rahman. I have your blood results back from last week. How have you been feeling?
PT
Not great, doctor. Still very tired all the time, and I have been getting short of breath going up the stairs.
DR
I see. Your haemoglobin has come back at 9.8, which is below the normal range. Your ferritin is also low at 8. This confirms iron deficiency anaemia.
PT
Is that serious?
DR
It explains your symptoms. I am going to start you on ferrous sulphate 200mg twice daily, and we should investigate the cause. Given your age and symptoms, I would like to arrange a colonoscopy to rule out any bleeding in the bowel.
PT
Okay. When would that be?
DR
I will put the referral through today as urgent. You should hear within two weeks. In the meantime, start the iron tablets and we will recheck your bloods in six weeks.
AI Scribe · Real-Time Transcription

The consultation is the documentation. No typing required.

When a consultation begins, the AI Scribe activates and transcribes the dialogue in real time using medical-grade automatic speech recognition with speaker differentiation. The system distinguishes between clinician and patient speech, captures medical terminology with clinical accuracy, and maintains the conversational flow as structured text. The clinician conducts the consultation normally — no dictation, no templates, no keyboard. The AI Scribe handles the documentation layer entirely.

Speaker Differentiation

Clinician · Patient · Multi-Speaker

AI distinguishes between clinician and patient voices using speaker diarisation. Multi-speaker consultations (family members, interpreters, multidisciplinary meetings) handled with per-speaker attribution.

Medical-Grade ASR

Clinical Terminology · Drug Names · Dosages

Speech recognition trained on clinical vocabulary — drug names, dosages, anatomical terms, procedure names, diagnostic terminology. Medical accuracy, not generic transcription.

Real-Time Processing

Live Transcription During Consultation

Transcription appears in real time as the consultation progresses. The clinician can glance at the transcript during the consultation to verify key points are captured correctly.

Structured Note Generation

Conversation becomes clinical record. In under 60 seconds.

After the consultation ends, the AI Scribe transforms the raw transcript into a structured clinical note — SOAP format (Subjective, Objective, Assessment, Plan), discharge summaries, or referral letters depending on the consultation type. The system extracts clinical entities: symptoms, examination findings, diagnoses, medications prescribed, investigations ordered, follow-up plans, and safety-netting advice. The structured note appears within seconds of the consultation ending. The clinician reviews, edits if needed, and approves with a single click.

SOAP Format

Subjective · Objective · Assessment · Plan

The standard clinical note format used across primary and secondary care. Each section populated automatically from the consultation transcript with appropriate clinical language.

Multiple Output Formats

SOAP · Discharge · Referral · Letter

The same consultation transcript can generate multiple output formats — a SOAP note for the clinical record, a referral letter for the specialist, and a patient summary letter. Each format uses appropriate clinical language and structure.

Sub-60-Second Completion

Consultation to Signed Record

From the moment the consultation ends to a clinician-approved, signed clinical note in the patient record — under 60 seconds. The documentation burden that consumes 49% of physician time, eliminated.

AI-Generated SOAP Note
Pending Review
Subjective
58-year-old female presents with persistent fatigue and exertional dyspnoea on climbing stairs. Symptoms ongoing for approximately 4 weeks. No chest pain, palpitations, or weight loss reported.
Objective
Blood results reviewed: Haemoglobin 9.8 g/dL (low), Ferritin 8 µg/L (low), MCV 72 fL (low). Consistent with iron deficiency anaemia. Clinical examination: pallor noted. Cardiovascular and respiratory examination unremarkable.
Assessment
Iron deficiency anaemia — cause to be determined. Given age and presentation, lower GI investigation warranted to exclude colorectal pathology as source of occult blood loss.
Plan
1. Commence ferrous sulphate 200mg BD. 2. Urgent referral for colonoscopy via 2-week pathway. 3. Repeat FBC and ferritin in 6 weeks. 4. Safety-net: return if symptoms worsen, new symptoms develop, or rectal bleeding occurs.
Clinical Safeguards

Every note AI-generated. Every word clinician-approved.

Mandatory Review

No AI-generated note enters the patient record without explicit clinician review and approval. The clinician reads, edits if needed, and signs off before finalisation.

Edit Authority

Every section of the generated note is fully editable. The clinician retains complete authority to modify, correct, add, or remove any content before approval.

Transcript Retention

The original consultation transcript is retained alongside the structured note. Any discrepancy between transcript and note can be reviewed and resolved.

Audio Privacy

Consultation audio is processed for transcription only. Audio data is not stored after transcription unless explicitly configured by the institution. Patient consent workflow integrated.

<60s
Note Completion
From consultation end to clinician-approved, signed clinical note in the patient record.
4
Note Formats
SOAP notes, discharge summaries, referral letters, and patient summary letters — all from the same transcript.
2
Speaker Channels
Clinician and patient speech differentiated with per-speaker attribution and medical-grade ASR accuracy.
49%
Time Recovered
Average physician time currently spent on documentation — returned to direct patient care through AI Scribe automation.
Who Scheduling & AI Scribe Serves

Physicians & GPs

Clinicians conducting consultations with AI-automated documentation — eliminating the administrative burden of note-writing while maintaining full clinical control over the final record.

Reception & Admin

Administrative staff managing appointment bookings, patient scheduling, and clinic capacity through a dedicated reception dashboard with calendar views and notification management.

Practice Managers

Practice and clinic managers monitoring appointment utilisation, no-show rates, average consultation duration, and documentation completion times across the practice.

Advancing medical intelligence.
At the frontier of medicine.

AXV-18 · Telehealth

Secure encrypted video consultations.
Integrated with scheduling, records, and AI Scribe.

End-to-end encrypted video consultations between clinician and patient, fully integrated with the AXIVIS scheduling system, patient records, and AI Scribe documentation. The clinician conducts the remote consultation with the same clinical tools available in face-to-face encounters — access to patient history, concurrent laboratory results, imaging studies, and real-time AI transcription. No separate telehealth platform required.

End-to-End Encryption · TLS 1.3 · Scheduling Integration · AI Scribe Active · Patient Portal Access
Deployment Options

Deploy Telehealth standalone, or as part of the AXIVIS OS.

Standalone · AXV-13
Telehealth on its own.
Procure Telehealth as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Telehealth as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Video Consultation Interface

The same clinical environment. Regardless of physical location.

AXIVIS Telehealth is not a standalone video platform bolted onto the clinical workflow — it is embedded within it. When a scheduled telehealth appointment begins, the video consultation opens within the AXIVIS environment alongside the patient's clinical record, medication list, outstanding results, and prior consultation notes. The clinician sees the patient and their data simultaneously. The AI Scribe from AXV-04 activates automatically, transcribing the remote consultation in real time with the same speaker differentiation and SOAP note generation as in-person encounters.

Unified Clinical View

Video + Patient Record + Active Medications + Results

The telehealth interface displays the video feed alongside the patient's clinical context — no switching between windows, no separate logins, no context loss between video platform and clinical system.

AI Scribe Integration

AXV-04 Consultation Suite · Automatic Activation

When the telehealth consultation begins, AXV-04 AI Scribe activates automatically. The remote conversation is transcribed in real time, structured into SOAP format, and presented for clinician review at consultation end — identical to in-person workflow.

Screen Sharing

Share Imaging · Share Lab Results · Share Reports

Clinicians can share specific clinical data with the patient during the consultation — imaging studies, laboratory trends, genomic reports — enabling informed shared decision-making in real time.

Video Consultation · Encrypted
14:23
Mrs. M. Rahman · Patient
Dr. Okonkwo
🎤
📹
🖥
📝
📋
✕
Platform Integration · Active During Call
03
AXV-03 Scheduling
Appointment booked, patient notified, waiting room active, follow-up auto-scheduled
04
AXV-04 Consultation Suite
AI Scribe active, real-time transcription, SOAP note generation at consultation end
05
AXV-05 Patient Records
Full clinical history, medications, allergies, and outstanding results visible during call
07
AXV-07 Orders & Diagnostics
Investigations ordered during call — lab requests, imaging referrals, prescriptions
06
AXV-06 Physician Connect
Secure post-consultation messaging between clinician and patient
Cross-Module Integration

Telehealth is not a feature. It is a consultation mode.

AXIVIS Telehealth does not exist as a separate tool — it is a delivery mode for the full clinical platform. During a telehealth consultation, every module available in face-to-face encounters remains accessible. The clinician can order investigations through AXV-07, review imaging through AXV-09, check laboratory trends through AXV-08, and prescribe medications through AXV-04 — all without leaving the consultation. The patient experience mirrors in-person care: the clinician is attentive, not distracted by separate systems, because the clinical tools are embedded within the same interface as the video feed.

Patient Waiting Room

Virtual Check-In · Connection Test · Pre-Visit Summary

Patients join a virtual waiting room before the consultation begins. The system tests audio/video connectivity, displays the pre-visit questionnaire if configured, and notifies the clinician when the patient is ready.

Multi-Party Consultations

Family Members · Interpreters · MDT

Support for multi-party video consultations — family members joining from separate locations, interpreter services, and multidisciplinary team meetings with multiple clinicians and the patient simultaneously.

Patient Portal Access

iOS · Android · Web Browser

Patients access telehealth consultations through the AXIVIS Patient App (iOS/Android) or web browser. No software installation required for the patient. Link-based join with authentication.

Consultation Security

Every frame encrypted. Every word protected. Every session governed.

Remote clinical consultations carry the same confidentiality requirements as in-person encounters — and additional risks from data transmission over public networks. AXIVIS Telehealth applies end-to-end encryption to all video, audio, and screen-sharing streams using TLS 1.3 with certificate pinning. No consultation data is stored on third-party servers. Session recordings, if enabled, are encrypted and stored within the AXIVIS Vault under the same access controls as all clinical documents. Every consultation session is logged with participant identities, connection timestamps, duration, and encryption verification.

End-to-End Encryption

TLS 1.3 · Certificate Pinning · Forward Secrecy

All video, audio, and data streams encrypted end-to-end. No intermediary can access consultation content. Certificate pinning prevents man-in-the-middle attacks.

Session Governance

Authenticated Join · Session Logging · Consent Tracking

Both clinician and patient authenticate before joining. Patient consent for telehealth consultation recorded. Session metadata (not content) logged for governance: who joined, when, duration, connection quality.

No Third-Party Dependency

Native AXIVIS Infrastructure

Telehealth runs on AXIVIS infrastructure — no Zoom, no Teams, no third-party video platform. Clinical data never transits through external servers. The same compliance certifications (ISO 27001, SOC 2 Type II, HIPAA) cover telehealth as all other AXIVIS modules.

Session Security Status
Encryption
TLS 1.3 ACTIVE
Certificate
PINNED · VERIFIED
Patient Consent
RECORDED
AI Scribe
ACTIVE · TRANSCRIBING
Recording
DISABLED
Third-Party Servers
NONE
Clinical Safeguards

Every consultation encrypted. Every participant authenticated.

End-to-End Encryption

All video, audio, and screen-share streams encrypted with TLS 1.3 and certificate pinning. No intermediary access to consultation content at any point in transmission.

Patient Consent

Telehealth consent recorded before each consultation begins. Consent status documented in the session log and the patient record. Configurable per institution.

Session Logging

Participant identities, join/leave timestamps, connection duration, encryption status, and AI Scribe activation all logged per session. Available for governance review.

Native Infrastructure

No third-party video platforms. All telehealth traffic runs on AXIVIS infrastructure under the same ISO 27001, SOC 2 Type II, and HIPAA compliance certifications.

E2E
Encryption
End-to-end encrypted video, audio, and screen sharing with TLS 1.3 and certificate pinning.
5
Module Integrations
Scheduling, Consultation Suite, Patient Records, Orders, and Physician Connect — all active during telehealth.
3
Access Methods
iOS app, Android app, and web browser — no software installation required for patients.
0
Third-Party Servers
All telehealth infrastructure runs natively on AXIVIS — no Zoom, Teams, or external video platform dependency.
Who Telehealth Serves

Remote Consultations

Clinicians conducting follow-up appointments, medication reviews, results discussions, and triage assessments with patients who cannot attend in person — with full clinical tooling available throughout.

Rural & Underserved Access

Patients in geographically remote or underserved areas accessing specialist consultations, multidisciplinary team input, and ongoing care management without travel burden.

Multidisciplinary Teams

Multi-party clinical discussions combining physicians, specialists, patients, and family members in a single encrypted session with shared access to clinical data and imaging.

Advancing medical intelligence.
At the frontier of medicine.

AXV-17 · Cortex

Cross-departmental clinical AI.
Context-aware. Patient-specific. Always present.

AXIVIS Cortex is the institutional intelligence layer that operates across every module. It answers clinical queries, interprets imaging, reviews drug interactions, summarises literature, and predicts diagnostic trajectories — all within the context of the data currently on screen. Standalone mode for general clinical intelligence. Floating mode for page-specific context. Patient mode for dedicated per-patient reasoning. Every output advisory. Every response clinician-validated.

Institutional AI · Cross-Departmental · Context-Aware · Per-Patient Intelligence · LLM with Dynamic RAG
Deployment Options

Deploy Cortex standalone, or as part of the AXIVIS OS.

Standalone · AXV-06
Cortex on its own.
Procure Cortex as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Cortex as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Three Intelligence Modes

One AI. Three levels of clinical awareness.

Cortex adapts its context window based on where it is invoked — general clinical queries, page-specific intelligence, or deep patient-level reasoning.

01
Standalone Cortex
Full-page clinical intelligence interface. General medical queries, literature synthesis, drug interactions, imaging interpretation, oncology protocols, and genomic insights. Cross-departmental access to Radiology AI, Oncology, Genomics, Laboratory, Clinical Trials, and Orders.
02
Floating Contextual Cortex
Compact panel that floats on every page. Aware of the current screen context — the module being viewed, the data being displayed, the patient record open. Suggests context-specific queries: critical flags, pending reviews, oncology deltas, genomic alerts.
03
Patient-Dedicated Cortex
A dedicated Cortex instance per patient. Full access to that patient's complete clinical history — labs, imaging, genomics, medications, consultations, procedures. Answers questions about this specific patient. Predicts diagnostic trajectories. Suggests investigations.
Mode 01 · Standalone Cortex

Institutional medical intelligence. Ask anything.

Standalone Cortex is the full-page clinical AI interface. Clinicians ask any medical question and receive structured, evidence-referenced responses. The system accesses six capability domains simultaneously: Imaging Analysis (interpret DICOM findings and imaging differentials), Oncology Protocols (treatment lines, chemotherapy regimens, response criteria), Genomic Insights (variant classifications, mutation burden, targeted therapy), Lab Interpretation (biomarker panels, reference ranges, pathology analysis), Literature Synthesis (clinical evidence, trials, guidelines), and Drug Intelligence (interactions, dosing protocols, formulary review).

Six Capability Domains

Imaging · Oncology · Genomics · Labs · Literature · Drug Intelligence

Each domain represents a specialised reasoning module. Cortex routes queries to the appropriate domain and synthesises cross-domain responses when a question spans multiple areas.

Department Access

Radiology AI · Oncology · Genomic · Lab · Trials · Orders

Cortex has read access to data from all connected departments. A query about a specific imaging finding can reference laboratory context, genomic data, and oncology treatment history simultaneously.

Session Management

New Session · Export Transcript · Query History

Each Cortex session maintains conversational context. Clinicians can start new sessions, export full transcripts for documentation, and review query history for reference.

AXIVIS Cortex · Standalone Interface
MEDICAL AI
AXIVIS Cortex Standalone Interface
Cortex · Floating Panel · v3.2
CONTEXTUAL
AXIVIS Cortex Floating Contextual Panel
Mode 02 · Floating Contextual Cortex

Aware of what you are looking at. Before you ask.

The floating Cortex panel appears on every page within the AXIVIS platform. It is aware of the current context — which module is active, which patient record is open, which data is displayed on screen. When a clinician is reviewing a radiology report, Cortex suggests queries about the imaging findings. When viewing a genomic panel, it surfaces questions about variant significance. The suggestions are not generic — they are generated from the actual data currently visible. Department badges (RAD, ONC, GEN, LAB, TRIALS, ICU) indicate which cross-departmental intelligence streams are active.

Page-Level Awareness

Dynamic Context Injection

Cortex automatically ingests the data currently displayed on the active page. When a clinician asks a question, the AI has already loaded the relevant context — no copy-pasting data, no switching windows, no manual context provision.

Context-Specific Suggestions

Critical Flags · Pending Review · Oncology Delta · Genomic Alerts

Suggested queries adapt to the current view. On a patient overview, Cortex suggests reviewing critical flags. On an oncology timeline, it suggests examining treatment deltas. On a genomic report, it flags variant alerts requiring attention.

Voice & File Input

Speak or Attach Clinical Files

Beyond text queries, clinicians can use voice input for hands-free clinical questions or attach clinical files (DICOM, PDF, VCF, images) directly to the Cortex panel for immediate AI analysis.

Mode 03 · Patient-Dedicated Cortex

One Cortex per patient. Full clinical history loaded.

Every patient in AXIVIS has a dedicated Cortex instance. When a clinician opens a patient's record and activates Cortex, the AI loads the complete clinical dataset for that specific patient — every consultation note, every lab result, every imaging report, every genomic finding, every medication, every procedure. The clinician can then ask patient-specific questions: "What is this patient's creatinine trend over the last 6 months?", "Are there any drug interactions with the current medication list?", "Based on the genomic profile and tumour staging, which clinical trials is this patient eligible for?" Cortex reasons across the full longitudinal patient record, not just the data currently on screen.

Full Patient Context

Labs · Imaging · Genomics · Medications · Procedures · Notes

The patient-dedicated Cortex loads the entire clinical history. Longitudinal trends, cross-modal correlations, and temporal patterns are all available for AI reasoning — spanning years of clinical data in a single query.

Predictive Intelligence

Trajectory Modelling · Risk Prediction · Investigation Suggestion

Based on the patient's complete history, Cortex can model diagnostic trajectories — predicting likely progression, suggesting investigations that would resolve diagnostic uncertainty, and flagging risk factors that require monitoring.

Cross-Modal Reasoning

Imaging + Labs + Genomics + Medications → Unified Clinical Picture

Ask Cortex to correlate a rising tumour marker with the latest imaging findings and genomic profile. The AI synthesises across data types that would normally require manual cross-referencing across multiple systems and departments.

MR
Mrs. M. Rahman · DOB: 14/03/1968
MRN: 447821 · Female · 58 years
IRON DEFICIENCY
COLONOSCOPY PENDING
FERROUS SULPHATE 200MG
HBA1C 7.2%
eGFR 48
PATIENT CORTEX · FULL HISTORY LOADED · 847 DATA POINTS
What is the creatinine trend over the last 12 months and does it indicate progressive renal impairment?
Are there any drug interactions between ferrous sulphate, metformin, and the current medication list?
Given the iron deficiency and pending colonoscopy, what is the differential diagnosis for the underlying cause?
Based on HbA1c trajectory and current eGFR, what is the risk of diabetic nephropathy progression?
Summarise all outstanding investigations and pending referrals for this patient.
Ask about this patient...
SEND ›
Clinical Safeguards

Every response advisory. Every output clinician-reviewed.

Advisory Output Only

All Cortex responses are clinical decision support. Cortex does not perform autonomous diagnosis, treatment selection, or prescribing. Every output requires clinician interpretation and validation.

Refusal Protocol

Cortex refuses out-of-scope queries, flags insufficient data for reliable response, and declines unsafe or unsupported requests. The system will not generate responses beyond its validated capabilities.

Full Inference Logging

Every Cortex query logged with: input text, model version, full response, context window contents, timestamp, and the clinician identity. Complete audit trail from question to answer.

CE MDR Compliance

All Cortex outputs display the regulatory disclaimer: "AI-generated responses · CE MDR 2017/745 · Not for direct clinical decision." Visible on every response and every interface mode.

3
Intelligence Modes
Standalone, floating contextual, and patient-dedicated — each with appropriate context depth and clinical awareness.
6
Clinical Domains
Imaging, oncology, genomics, laboratory, literature synthesis, and drug intelligence — all queryable through natural language.
<5s
Response Time
Average time from clinical query to structured, evidence-referenced response with citation backing.
100%
Query Audit
Every query, every response, every context window logged immutably with clinician identity and model version.
Who Cortex Serves

All Clinicians

Every physician, specialist, and allied health professional using AXIVIS has access to Cortex — from general medical queries to patient-specific diagnostic reasoning, across every department and every module.

Multidisciplinary Teams

MDT discussions supported by Cortex cross-modal reasoning — correlating imaging, genomics, laboratory, and treatment data in real time during tumour board and clinical decision meetings.

Clinical Education

Trainees and junior clinicians using Cortex for evidence-based learning — literature synthesis, guideline retrieval, and clinical reasoning support with full citation backing.

Advancing medical intelligence.
At the frontier of medicine.

AXV-06 · Physician Connect

Secure clinical messaging.
Clinician-to-clinician. Clinician-to-patient.

Encrypted messaging between clinicians and between clinicians and patients. No AI processing on message content. Role-based access controls enforced at the thread level. Every message logged with sender identity, recipient identity, and timestamp. Clinical attachments supported — share lab results, imaging reports, and referral documents directly within the conversation thread.

End-to-End Encryption · No AI on Message Content · RBAC Enforced · Full Audit Trail
Deployment Options

Deploy Physician Connect standalone, or as part of the AXIVIS OS.

Standalone · AXV-14
Physician Connect on its own.
Procure Physician Connect as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Physician Connect as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Clinician-to-Clinician Messaging

Clinical communication that stays within the clinical system.

Clinical conversations about patient care happen constantly — between GPs and specialists, between radiologists and oncologists, between surgeons and anaesthetists. Currently, these conversations happen through email, WhatsApp, pager systems, and corridor conversations — none of which create a clinical audit trail, none of which are encrypted to healthcare standards, and none of which are linked to the patient record. AXIVIS Physician Connect provides a secure messaging channel that exists within the clinical environment, linked to patient records, encrypted end-to-end, and fully auditable.

Thread-Level RBAC

Role-Based Access · Patient-Linked Threads

Messaging threads linked to specific patient records inherit the access controls of that patient's data. Only clinicians with authorised access to the patient can participate in patient-linked threads.

Clinical Attachments

Lab Results · Imaging · Referrals · Documents

Share clinical documents directly within the messaging thread. Attached files inherit the same encryption and audit controls as all AXIVIS Vault documents. No need to download, re-upload, or use external file-sharing.

No AI Processing

Privacy by Design

Physician Connect is explicitly excluded from AI processing. Message content is not analysed, summarised, or processed by any AI model. Communication between clinicians remains private and human-only.

Messages
Dr. R. Harwood
CT report attached — 14mm RLL nodule, recommend PET
Today, 10:14
Dr. S. Mitchell
BRCA1 result confirmed pathogenic. MDT Thursday?
Today, 09:22
Mrs. M. Rahman
Thank you doctor. I will start the tablets today.
Yesterday
Dr. L. Chen
Colonoscopy slot available Friday 18th AM
Yesterday
Lab Team
Critical result — Creatinine 142, eGFR 48
Mon
RHDr. R. Harwood · Consultant Radiologist
Online
Good morning. I have reported the CT chest for your patient Mrs. Rahman. There is a 14mm spiculated nodule in the right lower lobe — new finding, not present on prior imaging.
10:08
I have attached the structured report. My recommendation is PET/CT for metabolic characterisation. Happy to discuss further.
10:09
Thank you Richard. I will review the report now and discuss with the patient this afternoon. Given the BRCA1 result, I think we should bring this to Thursday's MDT.
10:14
Agreed. I will prepare the imaging for MDT presentation. Shall I also pull the comparison from 2024?
10:16
Yes please. That would be very helpful. I will add it to the MDT agenda now.
10:18
Type a message...
SEND
MRMrs. M. Rahman · Patient Portal
Last seen: 2 hours ago
Good afternoon Mrs Rahman. Your blood results are back. I would like to start you on iron tablets and arrange a further investigation. I have attached a summary letter for your reference.
Yesterday, 15:32
Thank you doctor. I will start the tablets today. Is the investigation the colonoscopy you mentioned?
Yesterday, 16:14
Yes, that is correct. The referral has been submitted. You will receive a separate appointment letter within two weeks. Please do not hesitate to message me if you have any questions in the meantime.
Yesterday, 16:22
Thank you. I appreciate you explaining everything clearly.
Yesterday, 16:30
Messages to patients are delivered via the AXIVIS Patient App and Patient Portal
Clinician-to-Patient Messaging

Patients deserve clinical communication that does not require a phone call.

Patients frequently need to communicate with their clinicians between appointments — asking about medication side effects, confirming appointment details, requesting result updates, or reporting symptom changes. Currently, this requires phone calls to reception, callback waiting, and verbal communication that leaves no record. Physician Connect provides a secure messaging channel between clinician and patient, delivered through the AXIVIS Patient App and Patient Portal. Messages are encrypted, linked to the patient record, and create a documented communication trail.

Patient App Delivery

iOS · Android · Web Portal

Patient messages delivered via push notification to the AXIVIS Patient App (iOS/Android) or accessible through the Patient Web Portal. Patients respond at their convenience — no hold times, no callback slots.

Documented Communication

Every Message in the Patient Record

All clinician-patient messages become part of the patient record. Follow-up instructions, medication clarifications, and clinical advice are documented — not lost in phone conversations or informal channels.

Attachment Sharing

Summary Letters · Results · Instructions

Clinicians attach patient-facing documents — consultation summary letters, result explanations, medication instructions, pre-procedure preparation guides — directly within the conversation thread.

Clinical Safeguards

Every message encrypted. Every thread governed. No AI processing.

End-to-End Encryption

All messages encrypted with TLS 1.3 in transit and AES-256 at rest. No message content accessible to intermediaries or third parties at any point.

No AI Processing

Physician Connect is explicitly excluded from AI analysis. Message content is not processed, summarised, or used for training. Clinical communication remains private and human-only.

RBAC Enforcement

Thread access controlled by role-based permissions. Patient-linked threads restricted to clinicians with authorised access. Unauthorised access attempts logged and flagged.

Complete Audit Trail

Every message logged with sender identity, recipient identity, timestamp, read receipt, and thread context. Full communication history available for clinical governance review.

2
Communication Channels
Clinician-to-clinician and clinician-to-patient messaging — both encrypted, both auditable, both linked to patient records.
E2E
Encryption
End-to-end encryption on all messages with TLS 1.3 in transit and AES-256 at rest. No cleartext message storage.
0
AI Processing
Zero AI analysis on message content. Physician Connect is explicitly excluded from all AI processing pipelines.
100%
Audit Coverage
Every message, every read receipt, every attachment share logged immutably with full sender/recipient metadata.
Who Physician Connect Serves

Clinicians

Physicians, specialists, and allied health professionals communicating about patient care through secure, documented channels instead of email, WhatsApp, or corridor conversations.

Patients

Patients communicating with their clinical team between appointments — medication questions, result queries, symptom updates — through the Patient App without phone calls or reception queues.

Clinical Governance

Governance teams with visibility into clinical communication patterns, response times, unread message alerts, and compliance with documented communication standards.

Advancing medical intelligence.
At the frontier of medicine.

AXV-13 · Clinical Trial Intelligence

Automated patient-trial matching.
Registry-connected. Evidence-structured. Not AI-inferred.

Connectivity to global trial registries — ClinicalTrials.gov, EU Clinical Trials Register, and WHO ICTRP. Patient clinical data compared against published eligibility criteria using structured data matching. Ranked match lists generated with inclusion/exclusion criteria assessment. Pre-populated referral letters for eligible trials. No AI inference — structured comparison only.

Registry Connectivity · Structured Matching · Ranked Results · Pre-Populated Referrals · No AI Inference
Deployment Options

Deploy Clinical Trials standalone, or as part of the AXIVIS OS.

Standalone · AXV-15
Clinical Trials on its own.
Procure Clinical Trials as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Clinical Trials as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Registry Connectivity

Three global registries. One search interface.

Every major clinical trial registry queried simultaneously through a single patient-linked search.

US
ClinicalTrials.gov
US National Library of Medicine. 450,000+ registered studies globally. Interventional and observational trials.
EU
EU Clinical Trials Register
European Medicines Agency. EU/EEA clinical trials under the Clinical Trials Regulation. Phase I–IV interventional studies.
WHO
WHO ICTRP
World Health Organization International Clinical Trials Registry Platform. Meta-registry aggregating 17 national and regional registries worldwide.
Automated Patient-Trial Matching

The patient's data compared against every eligible trial. Systematically.

When a clinician initiates a trial search for a patient, AXIVIS extracts the patient's structured clinical data — diagnosis, staging, biomarkers, mutation status, prior treatments, performance status, organ function, and demographics — and compares it against the published inclusion and exclusion criteria of all active trials matching the clinical context. Each trial receives a match score based on how many eligibility criteria the patient meets, partially meets, or does not meet. The result is a ranked list of potentially eligible trials, ordered by match strength.

Structured Data Matching

Not AI Inference — Criteria Comparison

Patient-trial matching in AXIVIS uses structured data comparison against published eligibility criteria. It does not use AI to infer eligibility. Each criterion is evaluated individually against the patient's data — met, partially met, or unmet — with transparent reasoning for each assessment.

Ranked Match Lists

Match Score · Criteria Coverage · Geographic Proximity

Matched trials ranked by criteria coverage (percentage of inclusion criteria met), geographic proximity to the patient's location, trial phase, and recruitment status. Clinicians see the strongest matches first.

Cross-Module Data Access

AXV-11 Oncology · AXV-12 Genomics · AXV-08 Labs

Trial matching pulls patient data from across AXIVIS modules — tumour staging from Oncology Intelligence, mutation status from Genomic Intelligence, organ function from Labs & Screening. No manual data entry required.

Patient-Trial Match Results
Mrs. M. Rahman · 8 trials found
NCT05847621
KEYNOTE-B92: Pembrolizumab + Chemotherapy in Advanced NSCLC with BRCA Alterations
94%
Phase III
Recruiting
London, UK
BRCA1+
NCT06012894
Olaparib Maintenance in BRCA-Mutated Solid Tumours — OLYMPIA-EXT
89%
Phase II
Recruiting
Birmingham, UK
PARP Inhibitor
EU-2024-003891
Atezolizumab + Bevacizumab in BRCA-Altered Advanced Malignancies
72%
Phase II
Recruiting
Amsterdam, NL
IO + Anti-VEGF
NCT05293017
Sacituzumab Govitecan in BRCA-Associated Epithelial Cancers
68%
Phase I/II
Enrolling
Manchester, UK
ADC
NCT04975997
Novel CDK4/6 Inhibitor in HR+/HER2- with Genomic Instability
41%
Phase I
Recruiting
Boston, US
CDK4/6i
Eligibility Criteria Assessment · NCT05847621
Criterion
Patient Data
Status
Age ≥ 18
58 years
MET
BRCA1/2 Alteration
BRCA1 c.5266dupC — Pathogenic
MET
ECOG PS 0-1
ECOG 1 (recorded 2026-03-15)
MET
Adequate Renal Fn
eGFR 48 mL/min (threshold: ≥ 30)
MET
Adequate Hepatic Fn
ALT 28, Bilirubin 14 — within range
MET
No Prior PARP Inhibitor
No PARP inhibitor in treatment history
MET
Measurable Disease
14mm RLL nodule (CT 2026-03-30)
MET
Platelets ≥ 100
245 ×10⁹/L
MET
Haemoglobin ≥ 9.0
9.8 g/dL
MARGINAL
Eligibility Criteria Assessment

Every criterion evaluated. Every data point traced to source.

When a clinician selects a matched trial, AXIVIS displays the full eligibility criteria assessment — each inclusion and exclusion criterion listed alongside the patient's relevant data point, the source of that data (which module, which date, which report), and the assessment result (Met, Partially Met, Unmet). The clinician sees exactly why a trial was matched and which criteria require further evaluation. No black-box matching — every comparison is transparent and verifiable.

Inclusion & Exclusion Criteria

Each Criterion Individually Assessed

Inclusion criteria assessed for coverage. Exclusion criteria assessed for absence. Each criterion linked to the specific data point in the patient record that supports the assessment — traceable from criterion to source document.

Partial Match Flagging

Marginal Values · Missing Data · Temporal Gaps

Criteria that are met marginally (e.g., haemoglobin 9.8 against a threshold of 9.0) or where the patient data is outdated are flagged as Partial. The clinician determines whether the criterion is clinically met based on current patient status.

Pre-Populated Referral Letters

Structured Referral with Eligibility Summary

For eligible trials, AXIVIS generates a pre-populated referral letter containing patient demographics, diagnosis, key clinical parameters, eligibility assessment summary, and trial site contact details. The clinician reviews and sends.

Clinical Safeguards

Structured comparison. Not AI prediction. Every match clinician-verified.

No AI Inference

Patient-trial matching uses structured data comparison against published eligibility criteria. The system does not use AI to predict eligibility, estimate outcomes, or infer clinical suitability beyond the published criteria.

Clinician Decision

The match list is decision support. The treating clinician determines whether a trial is appropriate based on the full clinical picture, patient preferences, and their own clinical judgement.

Transparent Criteria

Every match assessment is fully transparent — each criterion, each patient data point, and each source document visible. No opaque scoring. No unexplained rankings.

Registry Source Verification

All trial data sourced directly from official registries (ClinicalTrials.gov, EU CTR, WHO ICTRP). Trial status, eligibility criteria, and site information updated from registry feeds.

3
Global Registries
ClinicalTrials.gov, EU Clinical Trials Register, and WHO ICTRP — queried simultaneously for comprehensive trial coverage.
450k+
Registered Studies
Access to over 450,000 registered clinical studies globally through combined registry connectivity.
0
AI Inference
Zero AI prediction in eligibility assessment. Structured data comparison only. Transparent, verifiable, and reproducible.
Auto
Referral Generation
Pre-populated referral letters generated automatically with patient data, eligibility summary, and trial site contact details.
Who Clinical Trial Intelligence Serves

Oncologists

Oncology teams identifying clinical trial options for patients who have exhausted standard treatment lines or who may benefit from novel therapeutic approaches matched to their molecular profile.

Clinical Trial Coordinators

Research coordinators managing trial recruitment, screening eligible patients, and coordinating referrals between treating clinicians and trial sites with pre-populated documentation.

Patients

Patients and families accessing information about potentially eligible trials through the Patient Portal, with clear explanations of what each trial involves and how to discuss options with their clinical team.

Advancing medical intelligence.
At the frontier of medicine.

AXV-14 · AXIVIS Vault

Secure encrypted clinical document storage.
Classified. Searchable. Governed.

Every clinical document stored with AES-256 encryption, automatically classified by sensitivity level using NLP, searchable through natural language queries, and governed by role-based access controls with complete audit logging. Unlimited storage per patient. Every access recorded. Every document retrievable.

AES-256 Encryption · NLP Sensitivity Classification · Natural Language Search · RBAC · Immutable Audit
Deployment Options

Deploy AXIVIS Vault standalone, or as part of the AXIVIS OS.

Standalone · AXV-05
AXIVIS Vault on its own.
Procure AXIVIS Vault as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy AXIVIS Vault as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Encryption Architecture

Five security layers. Zero compromise.

Clinical documents contain the most sensitive data in healthcare — genomic results, psychiatric assessments, HIV status, substance use records, safeguarding reports. AXIVIS Vault applies a multi-layer security architecture to every stored document. AES-256 encryption at rest, TLS 1.3 in transit, HSM-managed encryption keys, role-based access controls enforced at the document level, and immutable audit logging of every access event. No document is stored unencrypted. No access goes unlogged.

AES-256 Encryption at Rest

Military-Grade Document Encryption

Every document encrypted with AES-256 before storage. Encryption keys managed through Hardware Security Modules (HSM) — keys never exist in plaintext outside the HSM boundary.

TLS 1.3 in Transit

Certificate Pinning · Forward Secrecy

All document transfers encrypted with TLS 1.3 with certificate pinning and forward secrecy. No document transmitted in cleartext under any circumstance.

Role-Based Access Control

Document-Level Permissions

Access permissions enforced at the individual document level, not the folder level. Clinicians see only documents their role authorises. Sharing requires explicit permission grant with audit trail.

Security Architecture · Active
E
AES-256 Encryption at Rest
Every document encrypted before storage with HSM-managed keys
ACTIVE
T
TLS 1.3 in Transit
Certificate pinning with forward secrecy on all transfers
ACTIVE
K
HSM Key Management
Hardware Security Module — keys never exist in plaintext
ACTIVE
R
Role-Based Access Control
Document-level permissions with least-privilege enforcement
ACTIVE
A
Immutable Audit Logging
Every access, view, download, and share event recorded permanently
ACTIVE
Patient Document Vault
Document
Type
Date
Level
PDF
Genomic Report — BRCA1/2 Panel
Genetics
2026-03-28
HIGH
PDF
Psychiatric Assessment — PHQ-9
Mental Health
2026-03-15
HIGH
IMG
CT Chest Report — Dr. Harwood
Radiology
2026-03-10
MED
HL7
Complete Blood Count — Lab
Laboratory
2026-03-08
MED
PDF
Consultation Letter — Dr. Okonkwo
Correspondence
2026-02-22
LOW
PDF
Referral — Oncology MDT
Referral
2026-02-18
MED
VCF
WES Variant File — Somatic Panel
Genomics
2026-02-10
HIGH
NLP Sensitivity Classification

Every document classified. By content, not by folder.

When a document enters the Vault, NLP analyses the content and automatically classifies its sensitivity level — High, Medium, or Low. Genomic results, psychiatric assessments, HIV-related records, and safeguarding documents are classified as High sensitivity with restricted access controls. Diagnostic reports and laboratory results are classified as Medium. Administrative correspondence is classified as Low. Classification is based on document content analysis, not file type or folder location — a PDF containing genomic data receives the same High classification as a VCF file.

Content-Based Classification

High · Medium · Low Sensitivity Tiers

NLP scans document content for sensitivity markers — genetic identifiers, mental health terminology, substance use references, safeguarding language, sexual health content. Classification assigned based on content, not metadata.

Natural Language Search

Semantic Query · Cross-Document Retrieval

Clinicians search the Vault using natural language queries — "latest genomic report", "blood results from March", "psychiatric assessment PHQ-9". Semantic search returns relevant documents ranked by relevance without requiring exact filename or date knowledge.

Unlimited Storage

No Per-Patient Storage Cap

No artificial storage limits per patient. Every document — from a single-page consultation letter to a multi-gigabyte whole genome sequencing file — stored without capacity restriction.

Access Governance & Audit

Every access logged. Every share tracked. Every download recorded.

Document access in healthcare is as sensitive as the documents themselves. AXIVIS Vault logs every access event — who viewed a document, when, from which device, for how long. Document sharing between clinicians requires explicit permission grant with documented clinical justification. Shared access can be time-limited and revoked. Downloads are logged with the downloading clinician's identity and the clinical context. The complete access history for any document is available for governance review at any time.

Access Event Logging

View · Download · Share · Print · Export

Every interaction with a document recorded: views with duration, downloads with destination context, shares with recipient and justification, prints, and exports. No silent access.

Sharing Controls

Explicit Grant · Time-Limited · Revocable

Document sharing requires the sharer to specify the recipient, clinical justification, and access duration. Shared access can be revoked at any time. Expired shares become inaccessible automatically.

Patient Access Portal

Patient-Controlled Document Viewing

Patients access their own Vault through the patient portal. They can view documents shared with them, download their own records, and see which clinicians have accessed their documents — full transparency.

Access Audit Log · Genomic Report
Today 09:14
Document Viewed
Dr. S. Mitchell · Clinical Geneticist · Duration: 4m 12s
Today 09:18
Shared with Clinician
Shared to: Dr. K. Okonkwo · Reason: MDT review · Expires: 7 days
Today 10:03
Document Viewed
Dr. K. Okonkwo · Oncology Consultant · Duration: 2m 48s
Today 10:06
Document Downloaded
Dr. K. Okonkwo · Context: MDT preparation · Encrypted PDF
Mar 28
Document Uploaded · Auto-Classified HIGH
System · NLP detected: genetic identifiers, variant data, hereditary risk
Clinical Safeguards

Every document encrypted. Every access governed. Every action audited.

End-to-End Encryption

AES-256 at rest, TLS 1.3 in transit, HSM key management. Documents are never stored or transmitted in cleartext under any circumstance.

Automatic Classification

NLP analyses document content and assigns sensitivity level automatically. High-sensitivity documents receive restricted access controls without manual intervention.

Complete Audit Trail

Every view, download, share, print, and export logged with clinician identity, timestamp, device, duration, and clinical context. No access goes unrecorded.

Data Residency

Document storage configured to comply with country-level data residency requirements. Genomic and clinical data stored within the jurisdiction specified by the institution.

AES-256
Encryption Standard
Military-grade encryption applied to every document at rest with HSM-managed key infrastructure.
3
Sensitivity Tiers
High, Medium, and Low — automatically classified by NLP content analysis with role-based access enforcement.
∞
Storage per Patient
No artificial storage limits. Every document from consultation letters to whole genome files stored without capacity restriction.
100%
Access Audit Coverage
Every access event — view, download, share, print, export — logged immutably with full actor and context metadata.
Who AXIVIS Vault Serves

Clinicians

Physicians, specialists, and allied health professionals storing and retrieving clinical documents with natural language search and role-appropriate access to sensitive patient records.

Patients

Patients accessing their own document vault through the patient portal — viewing shared records, downloading personal health data, and monitoring which clinicians have accessed their documents.

Information Governance

Data protection officers and information governance teams monitoring access patterns, sensitivity classification accuracy, sharing compliance, and data residency adherence across the institution.

Advancing medical intelligence.
At the frontier of medicine.

AXV-04 · Consultation Suite

Clinical Decision Intelligence Engine.
Ambient capture. Structured reasoning. Actionable output.

Real-time multimodal clinical data synthesis with structured risk stratification, prescription safety validation, and referral pathway intelligence. From consultation audio to signed clinical record in under 60 seconds.

EU MDR Class IIa · ISO 13485 · Human-in-the-Loop Mandatory
Deployment Options

Deploy Consultation Suite standalone, or as part of the AXIVIS OS.

Standalone · AXV-03
Consultation Suite on its own.
Procure Consultation Suite as a single module. The module ships with its own worklist, reporting surfaces, analytics, and API. HL7 FHIR and DICOM interoperability out of the box — integrates with your existing EHR, PACS, and LIS without dependency on other AXIVIS modules.
Deployment48 hours
IntegrationHL7 FHIR · DICOM
LicensingPer seat or site
AXIVIS OS
Or part of the unified AXIVIS OS.
Deploy Consultation Suite as one module of the full AXIVIS Clinical Intelligence OS — alongside all 16 modules sharing a single patient record, unified Cortex context, cross-module flagging, and one integrated worklist. Module outputs flow into adjacent modules automatically — labs cross-flag into procedures, imaging findings propagate into oncology.
Modules16 unified
Patient RecordSingle longitudinal
Cortex ContextCross-module
See the AXIVIS OS → Request Standalone Pricing
Functional Architecture

Three-layer clinical intelligence pipeline.

Data ingestion, clinical reasoning, and structured output — each layer independent, each layer auditable.

Layer A · Input
Data Ingestion
Ambient audio capture (medical-grade ASR)
Speaker differentiation (clinician / patient)
DICOM imaging overlay
Laboratory values (structured + PDF)
Patient history + longitudinal records
Patient-reported symptom inputs
→
Layer B · Intelligence
Clinical Reasoning Engine
Pattern detection across modalities
Anomaly scoring (deviation from baseline)
Progression tracking (longitudinal delta)
Guideline alignment (evidence-based logic)
Risk classification (Low / Moderate / High)
Confidence scoring per output
→
Layer C · Output
Structured Clinical Output
Clinical Summary (SOAP / discharge / referral)
Risk Profile (stratified, confidence-scored)
Recommended Actions (investigation pathway)
Prescription Support (safety-validated)
Referral Recommendation Engine
Audit-ready documentation trail
Live Intelligence Pipeline

System processing sequence.

Real-time clinical reasoning from data ingestion to structured output.

AXV-04 Processing
Initializing
Phase 01
Data Ingestion
01
Ambient audio captured. Speaker differentiation active. DICOM CT chest loaded. Laboratory panel (CBC, CMP, LFT) ingested. Patient longitudinal record linked. All inputs validated and normalized.
Phase 02
Correlation
02
Cross-referencing imaging findings with laboratory markers. Identifying deviation from prior baseline values. Detecting pattern: elevated LDH + new pulmonary nodule + declining eGFR. Flagging for multi-parameter anomaly assessment.
Phase 03
Risk Evaluation
03
Risk classification engine active. Multi-parameter synthesis complete. Confidence score: 0.87. Risk level: MODERATE-HIGH. Progression indicator detected from longitudinal comparison. Escalation threshold met for specialist referral pathway.
Phase 04
Output Generation
04
Generating structured clinical summary. Risk profile compiled. Investigation pathway defined. Prescription safety checks complete. Referral recommendation engine activated. All outputs pending clinician review and approval.
Structured Output · Clinician Review Required
Clinical Summary
62M presenting with progressive dyspnoea. CT chest: 14mm spiculated RLL nodule, new since prior study. Laboratory: LDH elevated 280 U/L, eGFR declining 58→52 over 3 months. Clinical correlation with weight loss and fatigue reported.
Risk Profile
Classification: MODERATE-HIGH
Confidence: 0.87
Basis: New pulmonary nodule + elevated LDH + declining renal function + constitutional symptoms. Multi-parameter anomaly cluster detected.
Recommended Actions
1. PET/CT staging study
2. CT-guided biopsy if PET-avid
3. Repeat renal function 2 weeks
4. Pulmonology referral for bronchoscopy assessment
Referral Engine
Specialist escalation triggered: Respiratory Medicine (primary), Oncology (conditional on histology). Urgency: 2-week pathway. Referral letter auto-drafted pending clinician approval.
Prescription Safety
Metformin 500mg BD: ✓ No interactions. ⚠ Renal function declining — recommend monitoring eGFR at next review. Threshold for dose adjustment: eGFR <45.
Amlodipine 5mg OD: ✓ No interactions detected. No contraindications identified.
Ambient Clinical Capture

Medical-grade speech recognition. Zero manual activation.

The consultation proceeds naturally. AXIVIS captures the full clinical dialogue using medical-grade automatic speech recognition tuned to clinical vocabulary. Speaker differentiation separates clinician from patient. The system structures the transcript into clinical note format in real time — SOAP, discharge summary, referral letter, or institutional custom template. No dictation. No templating. No retrospective documentation.

Speaker Differentiation

Real-Time Separation

Clinician and patient dialogue separated automatically. Clinical assertions distinguished from patient-reported symptoms.

Structured Note Generation

SOAP · Discharge · Referral · Custom

AI composes complete clinical documents — not raw transcription but intelligently structured clinical records.

Diagnosis Extraction

AI-Assisted Identification

Potential diagnostic terms extracted from transcript and presented for explicit clinician confirmation before entry into record.

Recording Active
03:42
DR
The chest CT shows a 14-millimetre spiculated nodule in the right lower lobe that was not present on the prior study from March.
PT
I have been having more shortness of breath over the last few weeks, and I have lost about four kilograms without trying.
DR
Given the new nodule and your symptoms, I want to arrange a PET scan and refer you to the respiratory team for further assessment.
Prescription Safety Engine
Warfarin 5mg
OD
Amiodarone 200mg
OD
Metformin 500mg
BD
CRITICAL: Amiodarone potentiates warfarin via CYP2C9/3A4 inhibition. Reduce warfarin 30-50%. Close INR monitoring required.
ADVISORY: eGFR declining (52 mL/min). Monitor renal function — metformin dose adjustment threshold eGFR <45.
Amiodarone-Metformin: No clinically significant pharmacokinetic interaction identified.
Prescription Intelligence

Multi-parameter drug safety validation at the point of prescribing.

Every prescription is validated against the patient's complete medication list, allergy record, renal function, hepatic function, pregnancy status, age, and weight. Drug-drug interactions, duplicate therapy, contraindications, and dose-adjustment requirements are evaluated simultaneously. Critical interactions block prescription completion pending documented clinician override. Advisory alerts inform without interrupting workflow.

Critical Interaction Block

Non-Bypassable Without Documented Rationale

High-risk prescribing errors blocked at the system level. Clinician must provide documented clinical rationale to proceed.

Pharmacogenomic Cross-Reference

When Genomic Data Available

If the patient has genomic data in AXIVIS, prescription safety checks include pharmacogenomic interactions affecting drug metabolism and response.

Clinical Safeguards

Every output auditable. Every reasoning step traceable.

Advisory Only

All outputs are decision support. No autonomous diagnosis. No autonomous treatment selection. Clinician review and approval mandatory before record entry or clinical action.

Confidence Scoring

Every output includes a confidence score derived from data completeness, pattern strength, and evidence alignment. Low-confidence outputs trigger explicit uncertainty flags.

Refusal Protocol

Insufficient data for reliable inference triggers refusal — the system withholds output and explicitly flags data gaps rather than generating unreliable guidance.

Immutable Audit Trail

Every AI inference logged with model version, input-output linkage, reviewing clinician identity, timestamp, and patient record association. Exportable for regulatory submission.

<60s
Consultation to Record
From consultation end to completed, signed clinical record entry via the ambient AI workflow.
Real-time
Processing Latency
AI consultation outputs generated in real time during active consultations. Predictive monitoring operates continuously.
4
Simultaneous Modalities
Audio, imaging, laboratory, and clinical documentation processed simultaneously within a unified pipeline.

Advancing medical intelligence.
At the frontier of medicine.

Platform Overview · OS + Standalone Modules

One OS. Sixteen modules.Deployed your way.

AXIVIS is a Clinical Intelligence Operating System comprised of sixteen modules covering the full diagnostic, clinical, and administrative surface of modern medicine. Every module is fully functional on its own — procurement is à la carte, deployment is 48 hours, integration is HL7 FHIR and DICOM native. Or deploy them together as the unified AXIVIS OS, where modules share a single longitudinal patient record, one Cortex context, and cross-module intelligence flow out of the box.

Modules16
Deployment48 hours
ProcurementÀ la carte or OS
InteropHL7 FHIR · DICOM
Two Procurement Paths

Start with a single module. Or deploy the unified OS.

Every AXIVIS module is a standalone institutional product. The same engineering, the same regulatory positioning, the same EU MDR classification. You can procure one module at a time, or deploy the full OS — the decision is commercial, not technical.

Path 01 · Standalone Module
One module. One procurement. Live in 48 hours.
Procure a single AXIVIS module — Radiology AI, Labs & Screening, Oncology Intelligence, Procedure Intelligence, or any of the sixteen — as a standalone product. Each ships with its own worklist, report surfaces, analytics, governance dashboard, and public API. Integrates with your existing EHR, PACS, LIS, or EPR via HL7 FHIR and DICOM. No dependency on other AXIVIS modules.
  • Independent deployment — lives alongside your existing clinical IT stack
  • Own governance surface — per-module analytics, workload, and override tracking
  • Native EHR/PACS/LIS integration — HL7 FHIR, DICOM, QIDO-RS, WADO-RS, STOW-RS
  • Per-seat or site licensing — commercial terms sized to the deployment
  • Upgrade path — add adjacent modules later with zero migration friction
Deployment
48 hours
Licensing
Per seat · Site
Scope
1 module
Path 02 · AXIVIS OS
Sixteen modules. One operating system. One patient record.
Deploy AXIVIS as a unified Clinical Intelligence Operating System — all sixteen modules sharing a single longitudinal patient record, one Cortex reasoning context, one worklist surface, and cross-module intelligence propagation. Findings from Labs cross-flag into Procedure Intelligence. Imaging AI outputs propagate into Oncology. Pathology correlates with Genomics. The OS is the integration layer the hospital doesn't have to build.
  • Single longitudinal patient record across every module and data type
  • Unified Cortex context — one reasoning surface sees imaging, labs, genomics, procedures, and notes
  • Cross-module flagging — critical labs surface as pre-procedural advisories, imaging findings propagate into oncology
  • One worklist across radiology, laboratory, procedures, and clinical review
  • Enterprise licensing — deployment contracts sized to the institution
Deployment
48 hours
Licensing
Enterprise
Scope
16 modules
Sixteen Modules

Every module: standalone or OS.

Every card below is deployable as a standalone product or as part of the unified AXIVIS OS. Click any module to see the shipping product and request standalone pricing.

AXV-01
Medical Intelligence
Clinical knowledge retrieval and structured evidence synthesis at the point of care.
StandaloneOS
→
AXV-02
Oncology Intelligence
RECIST 1.1 lesion tracking, protocol adherence, CTCAE toxicity, and MDT orchestration.
StandaloneOS
→
AXV-03
Consultation Suite
AI-assisted clinical documentation, SOAP notes, and ambient consultation capture.
StandaloneOS
→
AXV-04
Genomic Intelligence
Variant interpretation with clinical significance, pharmacogenomics, and hereditary risk.
StandaloneOS
→
AXV-05
AXIVIS Vault
Encrypted institutional document vault with sensitivity classification and audit trail.
StandaloneOS
→
AXV-06
Cortex
Clinical reasoning engine with intent classification and safety-first response hierarchy.
StandaloneOS
→
AXV-07
Orders & Diagnostics
Unified diagnostic order lifecycle — labs, imaging, pathology, genomics, and procedures.
StandaloneOS
→
AXV-08
Labs & Screening
Laboratory report extraction, critical value alerting, and longitudinal trend tracking.
StandaloneOS
→
AXV-09
Radiology AI
CNN-based diagnostic imaging interpretation with structured reporting and key findings.
StandaloneOS
→
AXV-10
Procedure Intelligence
Frame-level CNN analysis of endoscopic, bronchoscopic, and interventional procedural video.
StandaloneOS
→
AXV-11
Scheduling & AI Scribe
Multi-physician scheduling with ambient AI scribe for consultation documentation.
StandaloneOS
→
AXV-12
Patients Portal
Patient-facing access to records, results, and secure clinician messaging.
StandaloneOS
→
AXV-13
Telehealth
Clinical video consultation with ambient AI documentation and integrated imaging.
StandaloneOS
→
AXV-14
Physician Connect
Clinician-to-clinician messaging with referral and patient-record attachment.
StandaloneOS
→
AXV-15
Clinical Trials
Automated patient-trial matching against ClinicalTrials.gov, EU CTR, and WHO ICTRP.
StandaloneOS
→
AXV-16
AXIVIS Imaging Suite
Zero-footprint browser-delivered DICOM workstation with twelve measurement primitives.
StandaloneOS
→
What Changes Between Paths

Every module works the same way in isolation or inside the OS.

The table below is not a feature-cut — every standalone module is fully functional. The OS adds intelligence propagation between modules. Nothing is hidden behind the OS that isn't in the standalone product.

Capability
Standalone Module
AXIVIS OS
Module functionality
● FullComplete module feature set
● FullComplete module feature set
Deployment time
● 48 hoursTo live clinical use
● 48 hoursPer module, parallel rollout
HL7 FHIR / DICOM interop
● NativeOut of the box, no middleware
● NativeOut of the box, no middleware
Patient record scope
Module-scoped — integrates with your EHR
● LongitudinalSingle record across all 16 modules
Cortex reasoning scope
Module-bound context — reasons over module data
● Cross-moduleReasons over imaging + labs + genomics + notes
Cross-module flagging
EHR-mediated via HL7
● NativeLabs → Procedures, Imaging → Oncology, etc.
Unified worklist
Module worklist only
● UnifiedOne worklist across all modules
Governance surface
● Per-moduleModule analytics and override tracking
● Department-wideAnalytics roll up across modules
Regulatory positioning
● EU MDR Class IIaWhere applicable to the module
● EU MDR Class IIaSame classification, unified deployment
Licensing
Per seat or per site
Enterprise, institution-wide
Integration Guarantees

What you get either way.

Whether you deploy one module or all sixteen, these guarantees are invariant. They are commitments of the AXIVIS platform architecture, not premium features.

48-Hour Deployment
Any module deployable to live clinical use within 48 hours of commercial agreement. Infrastructure provisioned, integrations configured, clinical pilots started — same week.
Native Interoperability
HL7 FHIR R4, HL7 v2, and DICOM Part 10 support are built into every module. No middleware layer. No integration consultancy dependency. QIDO-RS · WADO-RS · STOW-RS for imaging modules; FHIR Bundle / Observation / DiagnosticReport for clinical modules.
Data Sovereignty
Configurable data residency region — EU, UK, US, APAC — per deployment. Encrypted at rest, encrypted in transit, institutional key management. Your data never leaves the region you specify.
Upgrade Path
Start with one module, add adjacent modules later without migration. The standalone module and the OS-deployed module are the same codebase. Adding a module means enabling it — not rebuilding.
Regulatory Alignment
EU MDR Class IIa classification where clinically applicable. ISO 13485 quality management system. SOC 2 Type II controls. HIPAA and GDPR alignment. Same regulatory posture whether standalone or OS.
Clinical Positioning
Every module is an augmentation layer for credentialed clinicians — not an autonomous diagnostic or therapeutic device. Every AI output is framed as preliminary and requires clinical validation. Human-in-the-loop is mandatory across every module.

Start with one module.
Or deploy the whole OS.

The commercial decision is yours. The technical architecture supports both. Book a procurement-grade walkthrough with the AXIVIS team and we'll scope the path that fits your institution — single module, phased rollout, or full OS deployment.

Solutions

Built for your
clinical environment.

AXIVIS delivers medical intelligence infrastructure built for the complexity of real clinical environments — from the independent physician to the enterprise health system.

Modular Procurement →
Deploy a single AXIVIS module standalone, or the full AXIVIS OS. Start with one module, add more later.
Individual Physicians →
A complete AI clinical workspace for the independent physician.
Small & Medium Practices →
Full clinical intelligence for growing medical practices.
Hospitals & Health Systems →
Enterprise intelligence infrastructure for health systems.
Healthcare Organisations →
Intelligence layer for clinics and integrated care organisations.
Universities & Academic Centres →
Infrastructure for academic medicine and clinical research.
Research Institutions →
Clinical data infrastructure for medical research and drug development.
Individual Physicians

A unified AI clinical workspace for the modern physician.

AXIVIS gives independent physicians the same diagnostic and clinical intelligence infrastructure used by major hospital systems — deployable from a single workspace without institutional IT overhead.

AI-assisted consultation documentation and SOAP notes
Radiology AI for imaging review and lab report interpretation
Prescription management with drug safety alerts
AXIVIS Vault for secure document storage
4–6 hrs
Recovered Daily
AI Scribe documentation, automated lab interpretation, order management, and prescription safety — administrative work that previously consumed the majority of the clinical day.
<60s
Per Consultation
From consultation end to clinician-approved, signed SOAP note in the patient record. No typing. No dictation. No templates.
18
Integrated Modules
Documentation, diagnostics, imaging AI, genomics, prescriptions, scheduling, lab analysis, orders, vault, and patient communication — unified in one clinical environment.
Small & Medium Practices

Clinical intelligence infrastructure for growing practices.

AXIVIS scales with your practice — from a single physician to a multi-specialty group — delivering unified diagnostic, clinical workflow, and patient intelligence without the complexity of enterprise deployment.

Multi-physician scheduling and appointment management
AI diagnostic tools across radiology, labs, and genomics
Shared patient records and clinical documentation
AXIVIS Vault for practice-wide document management

Role-Based

Access Controls

Clinical data accessible only to authorised users within the practice. Patient records protected at the appropriate access level for each staff role.

Multi-Clinician

Coordination

Shared diagnostic orders, patient records, and clinical communication across the full practice in one environment.

Audit-Ready

From Day One

Full audit logging of clinical actions, documentation edits, and record access. Compliance infrastructure built in, not added later.

Hospitals & Health Systems

Enterprise medical intelligence infrastructure for health systems at scale.

AXIVIS deploys as a unified intelligence layer across hospital departments — integrating diagnostic AI, oncology intelligence, genomic analysis, and clinical workflow tools within existing hospital infrastructure without replacing current systems.

Enterprise-grade diagnostic AI across all modalities
Oncology and genomic intelligence integration
Unified clinical documentation and workflow tools
Full institutional data sovereignty and residency
ISO 13485
Medical Device Quality
Compliance framework governing the quality management system underlying the AXIVIS platform.
SOC 2 Type II
Security Controls
Independent audit of security, availability, and confidentiality controls across the platform.
HIPAA & GDPR
Data Protection
Full alignment with international healthcare data protection requirements applicable across jurisdictions.
Full Audit Trail
Clinical Action Logging
Every documentation edit, record access, and clinical action logged immutably and available for review.
Healthcare Organisations

Medical intelligence infrastructure for organisations beyond the traditional hospital setting.

AXIVIS supports diagnostic clinics, specialist centres, imaging facilities, and integrated care organisations with the same AI intelligence layer deployed in major hospital systems.

Diagnostic order lifecycle management
AI radiology and lab interpretation
Secure document vault with sensitivity classification
Role-based access control across departments

Multi-Site

Deployment

Deploy AXIVIS infrastructure medicinally across multiple clinical sites with centralised governance and access control.

Coordinated

Patient Records

Patient clinical history accessible across the organisation, with role-based controls ensuring appropriate access at each site and process level.

Compliance

At Scale

ISO, SOC 2, HIPAA, and GDPR compliance frameworks applicable across the full organisational deployment.

Universities & Academic Medical Centres

Medical intelligence infrastructure for academic medicine and clinical research.

AXIVIS provides academic medical centres and universities with the clinical intelligence and research data infrastructure needed to advance medical AI research and clinical training programmes.

Clinical trial patient matching and eligibility screening
Real-world multimodal clinical datasets
Genomic and biomarker intelligence for research
AI diagnostic tools for clinical training

Clinical & Research

Data in One Infrastructure

Patient care data and research data managed within the same infrastructure, with appropriate access controls separating clinical and research functions.

Translational

Research Support

Genomic intelligence, diagnostic data, and clinical records available for translational research activities within the platform's compliance framework.

Academic

Governance Alignment

Compliance and audit infrastructure designed to meet the requirements of academic institutional governance and research ethics frameworks.

Research Institutions

Real-world clinical data infrastructure for medical research and drug development.

AXIVIS collaborates with research institutions and pharmaceutical organisations to build large-scale medical intelligence datasets across oncology, genomics, diagnostics, and precision medicine.

Multimodal real-world clinical datasets
Genomic and pharmacogenomic data pipelines
Oncology imaging and biomarker repositories
Clinical trial matching and recruitment infrastructure

Genomic

Research Infrastructure

VCF ingestion, variant annotation, pharmacogenomic modelling, and disease-risk infrastructure available for research workflows.

Longitudinal

Biomarker Data

Laboratory and biomarker data tracked longitudinally across patient records, accessible for research analysis within the compliance framework.

Clinical Trial

Infrastructure

Trial matching based on molecular and clinical criteria. Patient identification for potential trial enrolment supported within the platform.

Medical Specialties

AXIVIS intelligence modules built for the clinical problems of each medical specialty.

Oncology
Integrated cancer intelligence combining imaging, genomics, biomarkers.
Learn More →
Radiology
AI-assisted imaging interpretation across all modalities.
Learn More →
Pathology
Structured pathology report management and biomarker correlation.
Learn More →
Clinical Genetics
Genomic variant analysis and pharmacogenomic insights.
Learn More →
Gastroenterology
Procedural video AI and endoscopy intelligence.
Learn More →
Pulmonology
Radiology AI for chest imaging and pulmonary assessment.
Learn More →
Dermatology
AI-assisted skin lesion analysis and screening.
Learn More →
General Practice
Complete clinical workspace for GP consultations.
Learn More →
Modular Procurement

Start with one module. Scale to the full AXIVIS OS when you're ready.

AXIVIS is architected as sixteen standalone modules that also compose into a unified Clinical Intelligence Operating System. You do not have to procure the full OS on day one — start with a single module that addresses your most pressing clinical gap, prove value in 48 hours, and add adjacent modules later without any migration overhead. The standalone module and the OS-deployed module are the same codebase.

Sixteen standalone modules — each deployable independently in 48 hours
Per-module EU MDR Class IIa positioning where clinically applicable
HL7 FHIR · DICOM native interoperability with your existing EHR, PACS, LIS
Upgrade path to the full AXIVIS OS with zero migration friction
Per-seat, site, or enterprise licensing — commercial terms scaled to the deployment
16
Standalone Modules
Each one a fully functional institutional product with its own worklist, reporting surfaces, analytics, and public API. Procure one, procure several, or procure the full OS.
48 hrs
To Live Clinical Use
Any module deployable to active clinical operations within 48 hours of commercial agreement. Infrastructure provisioned, integrations configured, clinical pilots started within the same week.
Same
Codebase
The standalone module and the OS-deployed module are the same engineering. Adding a module to an existing standalone deployment means enabling it — not rebuilding.

Advancing medical intelligence.
At the frontier of medicine.

— AXIVIS CLINICAL INTELLIGENCE

ADVANCING MEDICAL INTELLIGENCE.
AT THE FRONTIER OF MEDICINE.

Clinical intelligence deployment records from live hospital environments. Each case documents AXIVIS modules, data flows, and outcomes across a specific specialty and patient pathway.

AXIVIS Black

Your complete biological picture.
Analyzed. Not just recorded.

Clinical-grade intelligence applied to your own biology. Every marker tracked. Every pattern surfaced. Every risk detected before symptoms appear.

HbA1c 5.2Glucose Marker
BRCA1 WTGenomic Status
PSA 0.8Biomarker
eGFR 92Renal Function
LDL 2.1Lipid Panel
CRP <1.0Inflammation
AXIVIS Black ingests your laboratory results, genetic data, imaging reports, and health records. AI analyzes them structurally, identifies risk patterns, tracks biological markers over time, and surfaces what requires attention. Not summaries. Not scores. Clinical-grade intelligence applied to your own data.
Most people have more health data than they can interpret. Most of it never connects.
Preventable chronic disease accounts for over 70% of global mortality. Most risk is detectable before symptoms appear.
Unlimited encrypted storage. Fully portable. Entirely yours.
WHAT AXIVIS BLACK DOES
01
Ingests Your Data
Laboratory results, genetic reports, imaging studies, medical records, and health documentation from any source or format. Everything in one place for the first time.
02
Analyzes Structurally
AI processes your data the same way a clinical system processes hospital patient data. Not summarized. Analyzed for patterns, markers, and biological deviations.
03
Maps Your Current Health State
Inflammation status, metabolic health, cardiovascular risk, hormonal balance, immune function. Structured from your actual data, not general population averages.
04
Identifies Cross-Signal Patterns
The Diagnostic Intelligence Hub consolidates your complete health data and flags patterns appearing across multiple data types simultaneously. Clinical insights that isolated systems miss.
05
Predicts and Surfaces Risk
Patterns in your biological data that indicate risk before symptoms appear. Prioritized by clinical significance, not algorithmic scores.
06
Routes You to the Right Care
Clinical triage intelligence analyzes your symptoms against your biological profile and health history. Identifies possible clinical patterns and directs you to the appropriate specialist.
07
Matches You to Clinical Trials
Your genetic profile, diagnosed conditions, and health history evaluated against available clinical trials. Relevant studies surfaced automatically.
08
Supports Your Mental Health
Continuous psychological wellbeing monitoring with AI-assisted support available around the clock. Mental health data integrated into your complete biological picture.

The problem is not a lack of data.
It is the absence of a system that connects it.

Preventable chronic disease is detectable years before diagnosis. The biological signals exist in laboratory data, genetic markers, and measurable biomarkers that most people have access to but no system to interpret at clinical depth.

74%
Global Deaths
Attributable to non-communicable diseases with detectable precursors.
5–10y
Lead Time
Measurable biological risk markers detectable years before diagnosis.
AXIVIS BIO-Score

Continuous biological monitoring. Not periodic checkups.

AXIVIS BIO-Score monitors your biological markers continuously as new data is added. Laboratory values, biomarker trends, and genomic risk factors are tracked longitudinally. When a pattern changes, you are informed before it becomes a clinical event.

01
Longitudinal Tracking
Every biomarker value tracked across your full data history.
02
Threshold Alert System
Approaching risk thresholds flagged before clinical values are reached.
03
Continuous Not Episodic
Monitoring updates every time new data is ingested.
Diagnostic Intelligence Hub

Every data signal connected. Patterns visible across your complete biological picture.

Individual health data points tell part of a story. A rising inflammatory marker means something different alongside an abnormal imaging finding and a genomic risk factor than it does in isolation. The Diagnostic Intelligence Hub consolidates your complete health data and identifies patterns appearing across multiple data types simultaneously.

The pattern is often more important than the individual result.

01
Cross-Signal Pattern Detection
AI identifies correlations across symptoms, biomarkers, imaging findings, and genomic data.
02
Consolidated Health View
Every data type in one structured environment. No switching between systems.
03
Continuous Pattern Monitoring
New data assessed against existing patterns every time it is added.
Biological Age Analysis

Your chronological age and your biological age are not the same number.

Biological age reflects how your body is functioning relative to population norms at the cellular and systemic level. Derived from genomic markers, inflammatory indicators, and metabolic function, AXIVIS Black generates a clinical-grade assessment of biological function.

01
Genomic Markers & Cellular Aging
Genetic variants associated with DNA repair mechanisms and cellular aging processes.
02
Multi-System Biological Assessment
Cardiovascular, metabolic, inflammatory, and immune markers analyzed together.
03
Personalized to Your Baseline
Assessments referenced against your own historical data, not just population averages.
Personal Genomic Analysis

Your genetic data interpreted at clinical depth.

Genetic test reports and raw sequencing data contain information about disease susceptibility, drug metabolism, and biological resilience. AXIVIS Black ingests your genetic data and generates a structured genetic health summary organized by clinical priority.

01
Disease Risk Stratification
Variants associated with elevated disease susceptibility identified and classified.
02
Pharmacogenomic Guidance
Genetic variants affecting drug metabolism and medication response identified.
03
Nutrition & Metabolic Factors
Genetic influences on nutrient metabolism and inflammatory response.
Clinical Triage Intelligence

Submit symptoms. Receive structured clinical direction.

AXIVIS Black analyzes your symptoms against your biological profile, health history, genomic data, and existing biomarker trends. The system identifies possible clinical patterns and directs you to the appropriate specialist.

The right specialist, identified from your data. Not from a list.

01
Contextualized Against Your Data
Symptom analysis conducted against your personal biological profile.
02
Possible Clinical Patterns
AI identifies clinically relevant patterns in your symptoms.
03
Specialist Routing
Direction to the appropriate clinical specialty based on symptom analysis.
PsycheCore

Psychological wellbeing monitored with the same rigor as physical health.

Mental health is a biological state. PsycheCore provides continuous psychological wellbeing monitoring and AI-assisted support available around the clock, integrated into your complete biological picture.

01
24/7 AI-Assisted Support
Psychological support and guidance available at any time. Available when you need it.
02
Integrated Biological Data
Mental health indicators tracked alongside physical biomarkers vs one system.
03
Continuous Monitoring
Psychological wellbeing assessed continuously as part of your overall health picture.
Clinical Trial Matching

Trials you may qualify for, identified from your actual profile.

AXIVIS Black evaluates your complete health and genetic profile against available clinical trials and surfaces studies you may be eligible for. Particularly relevant for individuals managing complex conditions.

01
Molecular & Clinical Criteria
Matching based on genetic variants, diagnosed conditions, and health history.
02
Updated Continuously
Trial database assessed each time your health profile updates.
03
Clinical Determination Remains Yours
Matching is informational. Final decisions remain with you and your team.
Health Data Vault

Every medical document you have ever generated. In one place. Searchable.

AXIVIS Health Vault stores all your records in one encrypted, searchable repository. NLP-based search retrieves any document using clinical language. Sensitivity classification protects the most private records.

01
NLP Search Across All Documents
Retrieve any document using natural language. Clinical context search.
02
Encrypted Personal Storage
All documents stored with encryption. Access controlled entirely by you.
03
Shareable When It Matters
Your complete medical history available to share with any clinician, anywhere.
When Your Data Indicates You Need More

AXIVIS Black connects directly to clinical care.

When your biological data surfaces findings that require professional evaluation, AXIVIS Black connects you to physicians and clinical services within the platform. Care appointments, physician messaging, and diagnostic orders are all available without leaving the environment where your data lives.

The data informs the consultation. The consultation informs the next action.

01
Direct Physician Connection
Connect with clinicians who have access to your complete biological picture.
02
Integrated Diagnostic Orders
Physicians can order follow-up diagnostics directly within the platform.
03
Data-Informed Consultations
Consultations driven by your actual data history, not just current symptoms.
WHO AXIVIS BLACK IS FOR

Individuals Managing Chronic Conditions

Structured biological monitoring, medication tracking, longitudinal data analysis, and clinical trial matching for people who need to understand their health data at the same depth their physician does.

Longevity & Precision Health

Biological age analysis, genomic risk stratification, continuous biomarker monitoring, and cross-signal pattern detection for individuals who approach health as a long-term scientific discipline.

Medical Professionals Personally

Physicians and clinical specialists who want the same intelligence layer they apply to patient data applied to their own. Clinical-grade analysis without the clinical context removal.

AXIVIS BLACK IS NOT
Not a fitness tracker.
Not a wellness score generator.
Not a generic symptom search engine.
Not a consumer health app.
AXIVIS Black applies clinical-grade AI to your personal health data. The same intelligence architecture that processes hospital patient records, radiology findings, and genomic sequencing data, applied to you as an individual. The output is not encouragement. It is clinical analysis.

Medical intelligence for individuals who take their biology seriously.

AXIVIS Black is currently available by request. Access is reviewed to ensure the platform is matched to the right individuals and use cases.

REQUEST ACCESS

AXIVIS Black access is reviewed individually. Tell us about yourself and how you intend to use the platform.
0 / 500
I agree to AXIVIS processing my information in accordance with the Privacy Policy.
RESEARCH INTELLIGENCE · AXIVIS CLINICAL DATASETS

BUILDING THE CLINICAL INTELLIGENCE
DATASETS THAT POWER
NEXT-GENERATION MEDICINE.

AXIVIS works with NHS trusts, academic medical centres, and international hospital networks to capture, structure, and annotate real-world clinical data at institutional scale. Every diagnostic event, genomic result, and treatment decision recorded through AXIVIS becomes a structured, research-ready intelligence asset — available under a rigorous, tiered governance framework.

Prospective and retrospective cohort data from live clinical deployments across oncology, radiology, genomics, ICU, neurology, and endocrinology.
All datasets clinician-validated at point of care — AXIVIS annotations are reviewed and approved by the treating clinician before entry into the research record.
GDPR, NHS DSP Toolkit, and ICH E6(R3) compliant data governance at every tier, with full provenance and de-identification audit trails.
18HOSPITAL
PARTNERS
10+COUNTRIES
REPRESENTED
4RESEARCH
DOMAINS
380KDICOM
STUDIES
142KANNOTATED
VARIANTS
24KLONGITUDINAL
CANCER CASES
RESEARCH DOMAINS
4 ACTIVE PROGRAMMES
01
CLINICAL AI DEVELOPMENT

Training and validation datasets for supervised and self-supervised learning across imaging, genomics, time-series vitals, and clinical text. Annotated for multi-task learning with clinician-approved ground truth labels at point of care.

DICOMNLPTIME-SERIESFOUNDATION
02
GENOMICS & PRECISION MEDICINE

Whole-exome and targeted panel VCF datasets linked to longitudinal clinical phenotypes, treatment response, pharmacogenomic profiles, and survival. CPIC-tiered variant annotations for every cohort with tumour-normal pairs for 4,200 solid tumour cases.

WESVCFPHARMACOGENOMICSCPIC
03
ONCOLOGY & LONGITUDINAL OUTCOMES

Structured cancer pathway records from primary diagnosis through MDT, treatment, response, and survivorship. Population-scale cohorts across 18 tumour types with 7-year median follow-up and linked biobank samples for 6,800 cases.

MDTSURVIVALREAL-WORLDBIOBANK
04
DIAGNOSTIC AI INFRASTRUCTURE

Multi-modal imaging corpora with radiologist-validated AI annotation overlays, procedural video with frame-level detection labels, and pathology slides linked to molecular classification and outcome data across 12 imaging modalities.

CT/MRIPETVIDEOHISTOLOGY

Every clinical interaction on the AXIVIS platform generates structured, annotated, research-ready data. What physicians use to treat patients becomes the intelligence that researchers use to advance medicine. No separate data collection pipeline. No retrospective annotation burden. No institutional silos.

AXIVIS RESEARCH INTELLIGENCE · INSTITUTIONAL FRAMEWORK 2026
FEATURED DATASETS
LIVE · UPDATED CONTINUOUSLY
AXIVIS · GENOMICSLIVE
GENOMICS · VCF · PHARMACOGENOMICS
High-fidelity variant data linked to clinical phenotypes

Whole-exome and targeted panel sequencing VCF datasets across oncology, rare disease, and pharmacogenomics cohorts. Every variant annotated at CPIC and ClinVar tier level and linked to longitudinal clinical outcome records. Includes matched tumour-normal pairs for 4,200 solid tumour cases and a pharmacogenomics sub-cohort of 3,100 patients with confirmed adverse drug event records.

142KVARIANTS
8,400PATIENTS
CPIC A–DTIER
TIER 2 ACADEMIC ACCESS
AXIVIS · ONCOLOGYLIVE
ONCOLOGY · LONGITUDINAL · MDT
Longitudinal cancer pathways at population scale

Generated through the AXIVIS Clinical Timeline, this dataset captures the entire journey from primary screening and pathology through MDT decision, treatment, response assessment, and long-term survivorship. Structured across 18 tumour types with 7-year median follow-up. Linked biobank samples available for 6,800 cases. MDT decision records include tumour board package content, treatment tier rationale, and clinical trial eligibility scores.

24KPATIENTS
18TUMOUR TYPES
7 YRFOLLOW-UP
TIER 3 INSTITUTIONAL ACCESS
AXIVIS · IMAGINGLIVE
IMAGING · DICOM · MULTI-MODAL
Multi-modal imaging findings with clinical correlation

Structural and functional imaging data across 12 modalities with DICOM annotations linked to lab values, genomic findings, and clinical outcomes. Each imaging episode cross-referenced to the full AXIVIS clinical record. Radiologist-validated AI annotation overlays for CT chest (iPE, nodule), CT head (stroke, ASPECTS 0–10), and MRI brain included. Procedure video sub-corpus of 14,000 laparoscopic cases with frame-level anomaly labels.

380KSTUDIES
12MODALITIES
AI+MDRANNOTATED
TIER 2 ACADEMIC ACCESS
DATA STANDARDS & METHODOLOGY
DATASET TYPECOLLECTION STANDARDANNOTATION METHODDE-IDENTIFICATIONUPDATE FREQUENCYACCESS TIER
CT / MRI ImagingDICOM 3.0 · HL7 FHIR R4Radiologist-validated AI overlay · Double-read for discordanceExpert Determination · Safe HarbourContinuousTIER 2
Genomic VCFVCF 4.2 · GA4GH · hg38CPIC tier annotation · ClinVar linkage · Manual review for pathogenick-anonymity k≥10 · Quasi-identifier suppressionQuarterlyTIER 3
Cancer PathwaySNOMED CT · ICD-10-CM · LOINCMDT-validated treatment records · Clinician-approved stagingDate shifting ±90 days · Age generalisationMonthlyTIER 3
Laboratory DataHL7 FHIR R4 · LOINC codesStructured result extraction · Reference range normalisationExpert DeterminationContinuousTIER 1
Procedural VideoMP4 H.264 · DICOM-SRFrame-level AI detection · Surgeon-reviewed bounding box labelsScene-level de-identification · Patient face suppressionMonthlyTIER 3
Clinical Notes (NLP)FHIR DocumentReference · HL7 C-CDAStructured concept extraction · NER entity labellingNLP-assisted PHI removal · Manual audit sampleQuarterlyTIER 4
PharmacogenomicCPIC Guidelines · PharmGKBVariant-drug pair annotation · Evidence-tiered prescribing flagsExpert Determination · Haplotype generalisationBi-annualTIER 2
ACCESS FRAMEWORK
4 TIERS · GOVERNED ACCESS
TIER 1 · OPEN
Aggregate & Benchmark

Anonymised summary statistics, model benchmarks, and validation metrics. Available without application for academic and commercial researchers.

  • Population-level statistics
  • Published benchmark datasets
  • Model performance reports
  • De-identified case counts
TIER 2 · ACADEMIC
Structured Clinical Datasets

Individual-level anonymised datasets for non-commercial academic research. Application-based with 4-week review by the Research Access Committee.

  • Imaging corpora (DICOM)
  • Pharmacogenomic VCF cohorts
  • Lab and outcome linkage
  • 12-month licence with renewal
TIER 3 · INSTITUTIONAL
Full Research Cohorts

Complete longitudinal cohorts including genomics, cancer pathways, and multi-modal imaging with clinical correlation. For NHS and academic institutions.

  • Genomic cohort linked records
  • Cancer pathway full records
  • Procedural video libraries
  • Secure compute environment
TIER 4 · COMMERCIAL
Derivative Model Licence

For commercial AI developers and pharmaceutical research divisions. Includes NLP datasets, proprietary annotation layers, and full API pipeline access.

  • NLP clinical note corpora
  • Proprietary annotation schemas
  • Full API pipeline access
  • Co-development agreements
HOW YOU CAN CONTRIBUTE
01
CLINICAL INSTITUTIONS

NHS trusts, private hospital groups, and specialist centres contribute anonymised clinical data under the AXIVIS Data Contribution Agreement. All governance managed within existing NHS IG and GDPR frameworks — no additional infrastructure required. Contribution partners receive access credit towards higher research tiers.

02
ACADEMIC MEDICAL CENTRES

University hospitals and academic departments contribute annotated research cohorts and participate in validation studies. Academic collaborators receive priority access to derived datasets, model benchmarks, and co-authorship rights on publications arising from contributed data under a structured Academic Collaboration Agreement.

03
RESEARCH INSTITUTIONS & BIOBANKS

Independent research institutes, biobanks, and pharmaceutical research divisions contribute validated datasets under a Data Transfer Agreement. All contributions are version-controlled and attribution-tracked within the governance audit trail. Biobank linkage is available for genomic cohort contributions meeting minimum size criteria of 500 patients.

PUBLICATIONS & BENCHMARKS
PEER REVIEWED
PREPRINT
Real-time pulmonary embolism detection on CT chest using multimodal clinical intelligence: A multicentre validation study across 18 hospital sites
AXIVIS Research Group · European Heart Journal Digital Health · Under review 2026
2026
PUBLISHED
Pharmacogenomic-guided prescribing at point-of-care: CYP2D6 and CYP2C19 integration in psychiatric inpatient settings — a prospective cohort study
AXIVIS Research Group et al. · The Lancet Digital Health · DOI: 10.1016/S2589-7500(26)00044-2
2026
BENCHMARK
AXIVIS NSCLC Recurrence Benchmark v2.1 — EGFR, KRAS, and ALK cohort detection performance across multicentre imaging and genomic data
AXIVIS Research Office · Available via Tier 1 Open Access · Updated quarterly
2025
PUBLISHED
Population-scale longitudinal cancer pathway modelling: Tumour board decision intelligence across 24,000 real-world cases
AXIVIS Research Group · Nature Medicine · DOI: 10.1038/s41591-025-03844-1
2025
DATASET
AXIVIS-STROKE v1.0 — Telehealth stroke triage dataset: 3,800 acute presentations with NIHSS, ASPECTS, thrombolysis decisions, and 90-day outcomes
AXIVIS Research Office · Zenodo · DOI: 10.5281/zenodo.14882234
2025
ETHICS & GOVERNANCE
REGULATORY FRAMEWORK

All AXIVIS research data collection operates under a single harmonised NHS Health Research Authority-approved protocol (REC ref: 25/NW/0142). Data processing activities are registered with the ICO under the AXIVIS Data Controller registration. All international data transfers comply with GDPR Chapter V adequacy and Standard Contractual Clauses requirements. The AXIVIS research programme holds MHRA Software as a Medical Device registration under the UK Medical Devices Regulations 2002.

DE-IDENTIFICATION PIPELINE

AXIVIS applies a three-stage de-identification pipeline: (1) automated PHI extraction using NLP entity recognition across all structured and unstructured fields; (2) quasi-identifier risk scoring using k-anonymity (k≥10) and l-diversity metrics per dataset release; (3) independent expert determination audit for all cohorts with cell counts below threshold. Full methodology is published in the AXIVIS Data Governance Specification v3.2, available on request.

PATIENT & PUBLIC INVOLVEMENT

AXIVIS operates a standing Patient and Public Involvement panel that reviews the research use framework annually and contributes to consent language and access policy design. All data use notifications are published in plain-language summaries accessible through the contributing hospital trust's patient portal. Patients retain the right to opt out of research data use at any time through the NHS Type 1 opt-out mechanism.

AUDIT & PROVENANCE

Every research data access event is logged in the AXIVIS Research Governance Ledger with full provenance: which dataset was accessed, by which institution, under which agreement, and for which stated research purpose. Annual audit reports are published publicly on the AXIVIS Research website. All access agreements include mandatory six-month reporting obligations and data destruction certificates on expiry, independently verified by the AXIVIS Data Protection Officer.

Apply for research data access.
Institutional, academic, and commercial access tiers available. Review typically 4 weeks.
Research → Intelligence

Medical Intelligence
Research

Advancing the science behind clinical AI reasoning.

AWAITING CONTENT
Research → Data

Real-World
Clinical Data

Structured, annotated datasets from real clinical practice.

AWAITING CONTENT
Research → Trials

Clinical Trials

Clinical trial infrastructure powered by real-world intelligence.

AWAITING CONTENT
COMPANY

BUILDING MEDICAL INTELLIGENCE
AT THE FRONTIER OF MEDICINE.

AXIVIS AI is a Medical AI and Clinical Research platform developed under Arenberg AG, a Swiss technology company headquartered in Zug, Switzerland.

ABOUT AXIVIS AI

The intelligence layer for
modern medicine.

AXIVIS AI is a Medical AI and Clinical Research platform built to transform fragmented diagnostic, genomic, and clinical data into real-time medical intelligence used by hospitals, physicians, research institutions, and individuals through AXIVIS Black.

AXIVIS AI operates as a dedicated Medical AI and Research Technology division under Arenberg AG, with its own clinical governance, compliance framework, and AI development infrastructure.

OUR MISSION

To accelerate the transition to proactive, intelligence-driven medicine by building the world's most advanced medical AI infrastructure.

OUR VISION

A future where clinical intelligence is ubiquitous, enabling every physician to deliver personalised, preventative, and precise care to every patient, globally.

AXIVIS · CLINICAL INTELLIGENCE NETWORK
2021FOUNDED
18HOSPITAL PARTNERS
10+COUNTRIES
ZugHEADQUARTERS
OUR STORY

How AXIVIS AI was built.

THE PROBLEM
Fragmented clinical data at scale.

Clinical data in modern medicine is highly advanced but fragmented across imaging systems, laboratory platforms, genomic databases, and clinical records. These systems operate in isolation, forcing clinicians to make decisions from incomplete information while large volumes of patient data remain unused.

THE GAP
Data exists. Access does not.

Even when data exists, it is not accessible in a usable form. Clinicians are required to interpret multiple disconnected outputs across systems, often under time constraints. The issue is not lack of data, but lack of integrated intelligence.

THE APPROACH
A unified intelligence layer.

The solution is not another standalone tool. It is a unified intelligence layer capable of ingesting multiple data types and delivering structured outputs directly within existing clinical workflows.

WHERE WE ARE NOW
Deployed across specialties.

AXIVIS AI operates as a unified platform across diagnostics, oncology, genomics, and clinical workflows, serving physicians, hospitals, research institutions, and individuals through AXIVIS Black.

AXIVIS · PLATFORM ARCHITECTURE
WHO WE ARE

Building Medical Intelligence Infrastructure.

We build medical intelligence infrastructure, not general purpose software. Every system is designed around a specific clinical problem, grounded in how physicians actually work across diagnostics, oncology, and genomics.

Our approach combines clinical research, real-world data, and AI engineering with strict standards for safety, traceability, and human oversight. Every output is structured, explainable, and designed to support clinical decision making, not replace it.

— Clinician-in-the-loop on every AI output
— Full audit trail and decision provenance
— Real-world validated, not benchmark-optimised
— NHS IG, GDPR, and MHRA compliant by design
COMPANY · ABOUT

About AXIVIS AI.

Medical AI and clinical research infrastructure built for institutional deployment.

WHAT WE BUILD

Medical intelligence infrastructure for the institutions that need it most.

AXIVIS is not a general-purpose AI tool. Every module is built around a specific clinical workflow — radiology triage, oncology pathway management, genomic variant interpretation, ICU surveillance — and is designed to operate within existing hospital infrastructure without requiring new hardware, separate login systems, or parallel documentation processes.

The platform ingests multimodal clinical data — DICOM imaging, genomic VCF files, laboratory results, procedure video, telehealth transcripts — and returns structured, clinician-reviewable intelligence outputs. No black boxes. No autonomous decisions. Every finding is traceable, auditable, and assigned to the responsible clinician.

CLINICAL DEPTH

Designed around how physicians actually work, not around what data science can produce.

HUMAN OVERSIGHT

Clinician review is required for every AI output. No autonomous clinical decisions.

REAL-WORLD DATA

Trained and validated on prospective clinical data, not curated academic benchmarks.

INSTITUTIONAL TRUST

NHS IG, GDPR, MHRA, and ISO-compliant. Built for enterprise healthcare environments.

LEADERSHIP

Guided by expertise.

AXIVIS is shaped by clinical experience, technical depth, and real-world medical workflows. Every feature is designed to support accurate and fast decision making in complex environments.

CLINICAL SCIENCE

Our clinical science team includes practising physicians from oncology, radiology, genomics, and emergency medicine who validate every module in real clinical environments before deployment.

AI ENGINEERING

Our AI and machine learning engineers specialise in medical imaging, genomic data processing, and clinical NLP. Every model is built to clinical-grade explainability and safety standards.

GOVERNANCE & COMPLIANCE

Our governance team maintains NHS IG, GDPR, and MHRA compliance across all product lines and research activities, with dedicated clinical safety officers for each platform deployment.

COMPANY · ARENBERG AG

Developed under
Arenberg AG.

A Swiss-based investment and development firm focused on advanced AI systems across healthcare, finance, and strategic intelligence.

STEWARDSHIP

Long-term commitment to clinical excellence.

AXIVIS AI is a core initiative of Arenberg AG, a Swiss-based investment and development firm focused on advanced AI systems across healthcare, finance, and strategic intelligence.

This stewardship ensures long-term stability, rigorous ethical standards, and a commitment to clinical excellence that transcends typical venture-backed cycles. Arenberg AG provides the governance infrastructure, capital stability, and institutional credibility that healthcare deployments require.

Arenberg AG is headquartered in Zug, Switzerland — a jurisdiction recognised for its rigorous financial regulation, data protection law, and technology governance standards, making it the natural home for a platform operating at the intersection of medical AI and institutional data.

VISIT ARENBERG AG ↗
WHAT ARENBERG BRINGS
STABILITY
Long-horizon capital.

Arenberg AG provides patient, long-horizon capital that insulates AXIVIS from short-term commercial pressures. Clinical AI requires years of validation, regulatory approval, and real-world refinement before it is ready for institutional scale.

GOVERNANCE
Swiss institutional standards.

Arenberg AG's Swiss incorporation brings rigorous governance frameworks, FINMA-adjacent regulatory discipline, and data protection standards aligned with Swiss-EU adequacy requirements — all of which flow through to AXIVIS AI's operational structure.

NETWORK
Cross-sector intelligence.

Arenberg AG's portfolio spans healthcare, finance, and strategic intelligence. This cross-sector perspective informs AXIVIS's approach to building infrastructure that meets the standards of regulated industries, not just healthcare software norms.

OVERSIGHT
Independent board oversight.

AXIVIS AI operates with independent board-level oversight through Arenberg AG, ensuring that product direction, clinical safety decisions, and research ethics are reviewed by a body with no short-term commercial incentive to accelerate beyond safe boundaries.

COMPANY · CONTACT

Let's build the
future together.

Whether you are a health system leader, a clinical specialist, or a potential partner, we'd love to hear from you.

SEND A MESSAGE
CONTACT DETAILS
✉
EMAIL
contact@axivis.ai
◎
HEADQUARTERS
Zug, Switzerland
◈
CLINICAL PARTNERSHIPS
partnerships@axivis.ai
⌬
RESEARCH ACCESS
research@axivis.ai
SUPPORT PORTAL

Existing customers can access technical support and documentation through the AXIVIS Support Portal.

🎧
Support Portal
Existing customers can access technical support and documentation.
SECURITY & COMPLIANCE

Enterprise-grade compliance.

AXIVIS AI platforms are built on internationally recognised compliance frameworks with governance structures that ensure transparency, traceability, and clinician oversight across all system layers.

COMPLIANCE FRAMEWORKS
ISO CERTIFICATIONS
ISO 9001:2015
QUALITY MANAGEMENT SYSTEMS
CERTIFIED
ISO/IEC 20000-1:2018
IT SERVICE MANAGEMENT
CERTIFIED
ISO/IEC 27001:2022
INFORMATION SECURITY MANAGEMENT
CERTIFIED
ISO/IEC 27701:2019
PRIVACY INFORMATION MANAGEMENT
CERTIFIED
ISO 13485:2016
PHYSICAL DEVICE QUALITY MANAGEMENT
IN PROGRESS
SOC COMPLIANCE
SOC 2 Type II

Rigorous independent auditing of security, availability, processing integrity, confidentiality, and privacy controls. Audited annually by a PCAOB-registered firm.

COMPLIANT
REGULATORY & DATA PROTECTION
GDPR (Aligned)
GENERAL DATA PROTECTION REGULATION
ALIGNED
HIPAA (Aligned)
HEALTH INSURANCE PORTABILITY AND ACCOUNTABILITY ACT
ALIGNED
CE Marking (EU MDR — In Progress)
EUROPEAN MEDICAL DEVICE REGULATION
IN PROGRESS
NHS DSP Toolkit
NHS DATA SECURITY & PROTECTION
COMPLIANT
SECURITY CONTROLS
INFRASTRUCTURE SECURITY
  • AES-256 encryption at rest
  • TLS 1.3 for all data in transit
  • Role-based access control (RBAC)
  • Multi-factor authentication (MFA)
  • Regular penetration testing
DATA SECURITY
  • Data minimisation — required data only
  • Purpose limitation — clinical use only
  • Full audit logs with decision provenance
  • Automatic session timeout
  • Zero standing access to production
AI GOVERNANCE
  • All AI outputs carry traceability logs
  • Clinician review required for all use
  • No autonomous clinical decisions
  • Human oversight at every layer
  • Confidence scoring on all outputs
PHYSICAL & OPERATIONAL
  • Swiss-based infrastructure (EU/CH)
  • Cloud, dedicated, and on-premise options
  • Business continuity & disaster recovery
  • Incident response procedures
  • Employee security training
DATA RESIDENCY & DEPLOYMENT
DEPLOYMENT OPTIONS
Cloud

Standard enterprise deployment with full compliance framework. Multi-tenant architecture with complete data isolation between institutions.

Dedicated Infrastructure

Dedicated infrastructure for institutions with data residency requirements. Single-tenant deployment on Swiss or EU-only infrastructure.

On-Premise

Full on-premise deployment for institutions with data sovereignty requirements. Full platform capability with local compute and storage.

DEPLOYMENT
Cloud, dedicated, and on-premise.

Cloud, dedicated infrastructure, and on-premise deployments for institutional requirements.

AI GOVERNANCE
Human oversight at every layer.

All AI outputs carry traceability logs. Clinician review required for all outputs. No autonomous clinical decisions.

DATA RESIDENCY
Swiss-based infrastructure.

Swiss-based infrastructure available for EU and GDPR-sensitive deployments. On-premise options for data sovereignty.

DATA PRIVACY

Privacy by design.

Patient data protection is not a compliance requirement for AXIVIS AI. It is a design principle.

Every system is built with privacy by design — data is encrypted, access is role-based, and every action on patient data is logged in a full audit trail.

AXIVIS AI does not use patient data for model training.
PRINCIPLES
  • Data minimisation
  • Purpose limitation
  • AES-256 encryption at rest and in transit
  • TLS 1.3 for all data transmission
  • Role-based access control with field-level permissions
  • Full audit logs with decision provenance
DOCUMENTATION & CERTIFICATES
CERTIFICATIONS
ISO 9001:2015 Certificate
ISO/IEC 20000-1:2018 Certificate
ISO/IEC 27001:2022 Certificate
ISO/IEC 27701:2019 Certificate
ISO 13485:2016 Certificate
SOC 2 Type II Letter of Attestation 🔒
FAQ
AXIVIS AI platforms are built and operated in accordance with ISO 13485:2016, ISO 27001:2022, ISO 9001:2015, ISO 20000-1:2018, and ISO 27701:2019. SOC 2 Type II independent auditing is maintained across security, availability, processing integrity, confidentiality, and privacy controls. The platform is aligned with HIPAA and GDPR requirements for the processing of protected health information and personal data respectively.
AXIVIS AI is pursuing CE Marking under the EU Medical Device Regulation (EU MDR 2017/745) and MHRA registration under UK Medical Devices Regulations 2002. ISO 13485:2016 certification is currently in progress. Existing deployments operate under the appropriate Software as a Medical Device (SaMD) classification for each module.
Patient data is processed within isolated, encrypted environments with zero data sharing between institutional tenants. All data is stored using AES-256 encryption at rest and transmitted using TLS 1.3. AXIVIS does not use patient data for model training or improvement. All processing activities are logged with full provenance, and data is stored exclusively on Swiss or EU infrastructure unless a specific deployment agreement specifies otherwise.
Access to patient data is governed by role-based access control (RBAC) with field-level permissions. Multi-factor authentication is mandatory for all clinical staff access. All access events — read, write, export, and deletion — are logged in an immutable audit trail with timestamp, user identity, and action type. Access logs are available to institutional administrators and can be exported for compliance reporting.
AXIVIS maintains 24/7 security monitoring with automated threat detection, anomaly identification, and incident response procedures. Regular penetration testing is conducted by independent third-party security firms. Vulnerability management follows a defined SLA with critical vulnerabilities patched within 24 hours and high-severity vulnerabilities within 72 hours.
Yes. AXIVIS AI is aligned with HIPAA Security Rule, Privacy Rule, and Breach Notification Rule requirements for the handling of Protected Health Information (PHI). Business Associate Agreements (BAAs) are available for US healthcare institutions and any institution processing data subject to HIPAA obligations under a contractual arrangement with US-based covered entities.
Yes. AXIVIS acts as a Data Processor under GDPR Article 28 for all institutional deployments. Data Processing Agreements (DPAs) aligned with GDPR requirements are available as standard. AXIVIS does not transfer data outside the EU/EEA or to non-adequate third countries without explicit contractual safeguards. A Data Protection Officer is available at dpo@axivis.ai for GDPR-related enquiries.
All AXIVIS AI models undergo prospective clinical validation in real hospital environments before deployment. Validation studies are conducted with participating NHS trusts and academic medical centres, with results reviewed by an independent Clinical Safety Officer. No model is deployed without clinician-in-the-loop confirmation of outputs, and confidence scoring is applied to all AI-generated findings to enable informed clinical review.
AXIVIS supports HL7 FHIR R4 for structured clinical data exchange and HL7 v2.x for legacy ADT, ORM, and ORU message processing. DICOM 3.0 is supported for all imaging data. Integration with NHS Spine, NHS MESH, and regional GPES systems is available for UK deployments. Full API documentation is available in the AXIVIS Technical Integration Guide.
Upon contract termination, AXIVIS provides a 90-day data export window during which the institution can retrieve all data in standard formats (FHIR R4, CSV, DICOM). Following the export window, all institutional data is securely deleted from AXIVIS systems within 30 days and a certificate of deletion is issued to the institution for compliance records.

Need compliance documentation
or have security questions?

Ready to deploy the future of medicine?

COMPLIANCE FRAMEWORKS ISO 9001:2015 ISO/IEC 20000-1:2018 ISO/IEC 27001:2022 ISO/IEC 27701:2019 ISO 13485:2016 SOC 2 Type II GDPR HIPAA CE Marking (EU MDR)
EXTERNAL VERIFICATION DATABASE
HOW TO VERIFY THIS CERTIFICATE
EXTERNAL VERIFICATION DATABASE
https://www.iafcertsearch.org ↗
01
Go to the website above
02
Enter the company name
e.g. Arenberg AG
03
Review the results for:
✓Certification Standard
✓Issuing Body
✓Validity Dates
✓Certificate Status
REQUEST ACCESS

SOC 2 Type II Letter of Attestation is available under NDA or to verified partner status.

FULL NAME (REQUIRED)
COMPANY NAME (REQUIRED)
WORK EMAIL (REQUIRED)
REASON FOR REQUEST

Book a Demo

Tell us about your practice or institution. We will configure the demonstration around your specific clinical workflows and data environment.

Full Name (Required)
Work Email (Required)
Organisation (Required)
Your Role
Organisation Type
Area of Interest
I agree to AXIVIS processing my information to arrange this demonstration. View Privacy Policy.
Get in touch
We'll respond within one business day.
ENQUIRY TYPE
FULL NAME *
WORK EMAIL *
ORGANISATION
ROLE / TITLE
MESSAGE
Your enquiry will be sent to support@axivis.ai
✓
Message sent.
Thank you for reaching out. A member of the AXIVIS team will be in touch shortly.
support@axivis.ai
AXIVIS Cortex
Clinical Intelligence Active

AXIVIS Cortex

Clinical-grade AI across oncology, genomics, radiology, and diagnostic intelligence. Ask me anything about AXIVIS.

All AI outputs are advisory and require clinician review. AXIVIS does not perform autonomous diagnosis.