AI identifies diabetes and blood pressure via facial videos

Tokyo University and the Tokyo Institute of Sciences have developed an artificial intelligence algorithm capable of diagnosing hypertension and diabetes by analyzing short video recordings of a person's face, without the need for contact. The study showed that the algorithm achieves over 90% accuracy in detecting hypertension and 88% for diabetes in 30-second video clips, and can be relied upon to expand early diagnostic screening. This contactless approach analyzes blood flow patterns and vascular flexibility by capturing images of the face and palms using a high-speed spectral camera. It aims to improve access to early diagnosis and reduce dependence on traditional tests that require clinical visits and direct measurements. The study anticipates that this method will aid in the early detection of serious conditions, thereby enhancing prevention of cardiovascular diseases. The results will be discussed at the European Society of Cardiology Conference 2026.

articleDetails.fontSize.sizeحجم النص

تنويه: هذا ملخص تم إنشاؤه بواسطة الذكاء الاصطناعي

عرض المقال بالكامل