AI advancements enable non-invasive detection of heart blockages and chronic diseases

Researchers are developing artificial intelligence tools to improve the detection of chronic health conditions through non-invasive methods.

At the Mayo Clinic, scientists have created an AI tool designed to identify significant heart blockages using standard ultrasound videos. The technology specifically targets individuals with hypertrophic cardiomyopathy (HCM), a condition that can cause abnormal thickening of the heart muscle. By analyzing B-mode images—standard videos that show the heart’s shape and motion—the AI can help identify obstructions in the left ventricular outflow tract (LVOT). This could assist clinicians in hospitals that lack specialized ultrasound experts.

Separately, researchers from the University of Tokyo and the Institute of Science Tokyo have developed a machine-learning algorithm capable of detecting hypertension and diabetes through facial video analysis. Using high-speed spectroscopic video of a participant's face and palms, the algorithm analyzes pulse-wave dynamics, skin blood-flow patterns, and spectral skin coloring. In studies, the algorithm achieved 95.0% accuracy in detecting hypertension from a 30-second recording and demonstrated high accuracy in detecting diabetes through facial blood flow patterns.