Project Overview
An NHS Teaching Hospital Trust processing 840,000 radiology examinations per year needed to implement AI clini...
Technology Stack
Compliance & Standards
The Challenge
An NHS Teaching Hospital Trust processing 840,000 radiology examinations per year needed to implement AI clinical decision support (CDS) for plain film X-ray and CT reporting — reducing radiologist reporting backlog (current 28,000 unreported examinations) whilst maintaining clinical safety. MHRA SaMD AI classification (Class IIa medical device — AI aiding diagnosis with moderate risk), IEC 62304 (software lifecycle for medical devices), DCB0129 (clinical safety — AI false positive/negative), NHS AI Lab Evidence Standards Framework, NHS DTAC all 5 domains, UK GDPR Article 9, and WCAG 2.1 AA were mandatory. Budget: £120,000.
Our Approach
MHRA AI classification
AI CDS that aids radiologist diagnosis = Software as a Medical Device (SaMD) Class IIa (MHRA MDR 2002 — UK MHRA post-Brexit classification).
Class IIa requirements
- 1UKCA marking (UK Conformity Assessed — CE mark no longer valid in GB from July 2024),
- 2UK Responsible Person registration with MHRA,
- 3Notified Body review (MHRA-approved UK Notified Body — for Class IIa, requires Notified Body certification),
- 4clinical performance data (sensitivity and specificity vs radiology gold standard — NHS DTAC requires peer-reviewed evidence),
- 5post-market surveillance (ongoing accuracy monitoring in production — MHRA requirement for AI medical devices).
AI CDS workflow
CT/X-ray image acquired → DICOM image sent to AI service (PACS/RIS integration via DICOM C-STORE) → AI model processes image → structured finding returned (anatomy, finding, confidence score, priority flag) → finding overlaid on PACS viewer (OHIF Viewer integration) → radiologist reviews AI finding + image → radiologist reports (AI finding is a prompt, not a decision).
Critical design
AI is a second reader that prompts, not a decision maker.
AI alert
- urgent finding flag (pneumothorax, PE — pulmonary embolism) → immediate radiology consultant alert (GOV.UK Notify).
- Radiologist can accept, modify, or reject AI finding — all decisions logged for audit.
- DCB0129 AI Clinical Safety —
AI clinical safety hazards
(1) false negative — AI misses significant pathology (radiologist over-relies on AI normal result → missed diagnosis).
Mitigation
display confidence score prominently, training for radiologists on AI limitations, audit of false negatives monthly. (2) False positive — AI flags non-existent pathology → unnecessary patient workup.
Mitigation
AI finding shown as "suggested" not "confirmed" — requires radiologist verification. (3) Demographic bias — AI performs differently across patient demographics (age, sex, ethnicity — documented in AI bias testing).
Mitigation
model validation across Trust's patient demographics, bias monitoring dashboard. (4) Distribution shift — CT scanner model changes → AI trained on old scanner → degraded performance.
Mitigation
AI performance monitoring (monthly accuracy check vs radiologist gold standard).
NHS AI Lab 14 Evidence Standards
- 1intended purpose clearly defined,
- 2training data documented (source, size, demographics),
- 3model architecture documented,
- 4performance metrics (sensitivity, specificity, AUC-ROC — vs radiologist gold standard),
- 5external validation dataset (different Trust population),
- 6bias testing (performance across demographic groups),
- 7deployment context documented (workflow, user training),
- 8monitoring plan (post-deployment performance tracking),
- 9human oversight mechanism (radiologist always final decision),
- 10explainability (attention maps — where AI is looking on image),
- 11safety validation (DCB0129 hazard log),
- 12benefit assessment (reduction in backlog × radiologist time saving),
- 13equity assessment (access for all patient groups),
- 14governance (NHS Trust Caldicott Guardian sign-off).
The Results
MHRA Class IIa UK Notified Body certification.
DTAC approved all 5 domains.
Platform live at 22 weeks, £112,000 — under budget.
Reporting backlog: 28,000 → 8,400 (70% reduction) in first 6 months.
AI sensitivity (plain X-ray): 94.2% (vs radiologist gold standard).
AI specificity: 96.8%.
False negative rate: 5.8% (industry benchmark 8–12% for equivalent models).
Radiologist time saving: 28% reduction in reporting time per examination.
Urgent finding alert: 100% of critical findings alerted within 5 minutes.
NHS AI Lab Evidence Standards: all 14 standards documented.
DCB0129: zero clinical safety incidents related to AI CDS in first 12 months.
“Reporting backlog from 28,000 to 8,400 — 70% reduction. AI sensitivity 94.2%, specificity 96.8%. False negative 5.8% versus 8–12% benchmark. Radiologist time 28% reduction per examination. Critical finding alert 100% within 5 minutes. NHS AI Lab all 14 standards. DCB0129 zero incidents. MHRA Class IIa UK Notified Body. The AI design principle — AI prompts, radiologist decides — was the clinical safety and MHRA regulatory framework that made adoption possible. Radiologists trust the system because it never overrides them." — Director of Radiology, NHS Teaching Hospital Trust (name withheld)”
Project Details
Related Case Studies
View AllHealthTech NHS Bowel Cancer Screening AI — FIT Result Analysis
An NHS Cancer Screening Programme laboratory processing 840,000 FIT (Faecal Immunochemical Test) samples per y...
HealthTech NHS Integrated Care Board Clinical Analytics Platform
An NHS Integrated Care Board (ICB) managing healthcare for 1.4M residents, £2.8B annual health expenditure, an...
HealthTech NHS Prostate Cancer Screening Digital Platform
NHS England piloting a targeted prostate cancer screening programme (Prostate Cancer UK — Transform trial and ...
HealthTech NHS AI Radiology Triage -- MHRA Class IIa
NHS Trust radiology processing 280,000 imaging studies per year needed AI-assisted triage prioritising urgent ...
Ready to Transform Your Business?
Let's discuss how our technical expertise can help you achieve remarkable results.