AI detects faces in 5 seconds for health screening
A recent Japanese study has revealed that artificial intelligence technology is capable of detecting indicators of diabetes and hypertension through short video recordings of the face and palms, lasting between 5 and 30 seconds. The study utilized an algorithm that analyzes blood flow, pulse dynamics, and skin tone characteristics, achieving over 90% accuracy in detecting high blood pressure and 88.2% in diagnosing diabetes when analyzing 30-second clips. When using shorter clips of 5 seconds, the accuracy was up to 81.2% and 90.3%, respectively. The technology is contactless and easy to use, representing a potential method for early detection of these diseases, which could lead to faster treatment initiation and improved public health outcomes. However, the current stage remains within research and evaluation.
Notice: This Is an AI-Generated Summary
Comments (0)