Clinical AI reports
Name of device or algorithm | Brief description | Data collection methods | FDA approval status | Type of algorithm | Data set composition | Population ethnic composition | Bias assessment evaluation | Model evaluation/Research protocol | Metrics for performance errors* † | Clinical workflow implementation |
| RadiologyIntel | Decision support software to augment medical imaging-related diagnosis | Standard H&E stained images, stimulated Raman histology | 510(k) Premarket notification | Convolutional neural network | Size/Composition of training dataset: 550 000 inpatients, academic medical centres Size/Composition of testing dataset: 350 000 inpatients at community hospitals | Non-Hispanic white 60% Hispanic and Latino 18% Black/African-American 13% Asian 6% Other 3% | Google TCAV Audit-AI | Multi-centred prospective clinical trial and retrospective analysis | Area under the curve 0.85 Classification accuracy 75% | Integrated into 50 hospitals via EHR systems, including Epic, Cerner |
| DiabetEYE | CDS system to enhance screening/diagnosis of diabetic retinopathy | Widefield stereoscopic photography and macular optical coherence tomography | De novo pathway | Convolutional neural network | Size/Composition of training dataset: 7000 outpatients, primary care clinic Size/Composition of testing dataset: 5000 Outpatients at independent clinic | Non-Hispanic white 70% Hispanic and Latino 10% Black/African-American 10% Other 10% | None available | Randomised controlled trial | Sensitivity, 81%, specificity, 90%, Area under the curve 0.80 Confusion matrix 0.91 | Implemented in 150 primary care clinics in the USA |
*Mishra.19
†Scott et al.20
AI, artificial intelligence.