Compare use cases
Drop two or three slugs into the URL — for example ?slugs=demo-mia-breast-screening,demo-idx-dr-retinopathy — or pick from the validated catalogue below.
Pick from validated:
AI imaging triage for the acute stroke pathwayAI second reader for breast cancer screeningAutonomous diabetic retinopathy screening in primary carePre-operative risk-stratification model for major surgerySelf-referral assistant for NHS Talking TherapiesType-2 diabetes long-term outcome predictor
Clear selection| Self-referral assistant for NHS Talking Therapies United Kingdom | Autonomous diabetic retinopathy screening in primary care Netherlands | Pre-operative risk-stratification model for major surgery Spain | |
|---|---|---|---|
| Trust badge | Validated | Validated | Validated |
| Evidence strength | Developing evidence | Robust evidence | Robust evidence |
| Pipeline stage | Validated | Validated | Validated |
| Intended use | Supports patient self-referral by collecting clinical information conversationally and routing referrals, reducing administrative wait time. | Provides an autonomous screening decision for more-than-mild diabetic retinopathy from retinal fundus images, referring positive cases to ophthalmology. | Computes a 30-day mortality and morbidity risk score from pre-operative records to support anaesthetic optimisation and shared decision-making. |
| Specialty | Psychiatry | Other | Surgery |
| Autonomy | Human in the loop (human acts) | Autonomous (acts without a human) | Informs a human (advisory) |
| Deployment | Active deployment | Active deployment | Active deployment |
| EU AI Act tier | Limited risk | High-risk | High-risk |
| EU MDR class | Not a device | Class IIb | Not a device |
| CE marking | Not required | CE marked | Not required |
| NICE ESF tier | Tier B — inform / monitor | Tier C — treat / diagnose / calculate risk | Tier C — treat / diagnose / calculate risk |
| Headline metric | Accuracy: 0.93 | Sensitivity: 0.87 | AUC / AUROC: 0.85 |
| Subgroup assessed | Yes | Yes | Yes |