Research programmes
Diagnostics researchStatus: research
Development-stage multimodal medical AI. Not a device, and not for clinical use.
Overview
- A development-stage research programme in multimodal diagnostic support, structured as six stages from ingestion through to deployment.
- Carries interpretability and uncertainty reporting as a named stage, because a diagnostic aid a clinician cannot interrogate is not usable.
- Not a medical device. No regulatory clearance, no published accuracy or sensitivity figures, and not for diagnostic use. Earlier published clinical numbers were withdrawn pending specialist review.
Part of AI & Automation.
How it works
Real components only. Anything not listed here does not exist yet.
01
Data ingestion
MRI, CT, PET, histopathology, genomics, EHR and biomarker sources
02
Feature encoding
Embeddings and kernels across the combined modalities
03
Hybrid modelling
Classification, feature extraction and prediction
04
Error mitigation
Noise and robustness controls
05
Interpretability
Confidence and uncertainty reporting alongside every output
06
Deployment
APIs and SDKs, cloud or on-premise
Specs
Actual limits, not aspirations. Where a row is missing, the number has not been measured or cleared for publication.
No specification is published for this product yet. We would rather show an empty table than numbers we cannot reproduce on request. Ask us directly and we will tell you what the current limits are.
Access
Research collaborations with clinical and academic partners. Not available as a product, and not for clinical use.