Seeing beyond BMI: Estimating cardiometabolic risk with smartphone imagery
Google says smartphone photos predicted insulin resistance nearly as well as DXA-derived body composition in its clinical validation.
PhotoScan estimates body fat percentage, android-to-gynoid fat ratio, and visceral-to-subcutaneous fat ratio from standard 2D phone images. In the independent MetabolicMosaic cohort, demographics plus PhotoScan reached an AUROC of 0.760, close to 0.773 for demographics plus DXA. Smartwatch BIA did not improve insulin-resistance classification in the same comparison. Google describes the system as an investigational research prototype, not a clinical screening product. Google Research's note
PhotoScan estimates body fat percentage, android-to-gynoid fat ratio, and visceral-to-subcutaneous fat ratio from standard 2D phone images. In the independent MetabolicMosaic cohort, demographics plus PhotoScan reached an AUROC of 0.760, close to 0.773 for demographics plus DXA. Smartwatch BIA did not improve insulin-resistance classification in the same comparison. Google describes the system as an investigational research prototype, not a clinical screening product. Google Research's note
score 5