
AI Diagnoses Acromegaly Skipping Facial Images
Kobe University researchers developed a privacy-first AI that detects acromegaly using only pictures of the back of the hand and a clenched fist, avoiding facial data. Trained on over 11,000 images from 725 patients across 15 centers in Japan, the model achieved high sensitivity and specificity and outperformed experienced endocrinologists, potentially speeding referrals and reducing healthcare disparities. The team plans to expand to other hand-visible conditions such as rheumatoid arthritis and anemia.





