For how we live.
Trained on 1.7 million ECGs paired with clinicians' report text, ECG-CLIP reached the same accuracy as the best comparator with about 90% less training data — a bid at the field's labelling bottleneck.
Stanford's SleepFM, published in Nature Medicine in January 2026, is the most consequential sleep paper of the year. Its own limitations section explains why it is not yet a clinical test.
Where AI clot-detection was tested head-to-head, it still missed more emboli than radiology trainees, especially small peripheral ones. Its clearest win was cutting time to diagnosis.
EyeFM was tested as an assistant to 16 ophthalmologists screening 668 high-risk patients in China. Patients in the AI arm also followed referral advice more often.
Off-the-shelf features from two published models matched retinal scans to the right patient 78% to 86% of the time. Fine-tuned for the task, one model reached 99.5% on OCT.