A new use-case of artificial intelligence or AI in smartphones might be able recognize stroke symptoms such as facial asymmetry

An innovative smartphone app that uses machine learning and AI to precisely identify the physical symptoms of strokes was disscused by researchers at the 20th Annual Meeting of the Society of Neurointerventional Surgery.

Researchers from the UCLA David Geffen School of Medicine and many medical facilities in Bulgaria participated in the study titled "Smartphone-Enabled AI Algorithms for Autonomous Stroke Detection." They gathered information from 240 stroke patients at four large stroke centers. The researchers used cell phones to capture videos of the patients and evaluate their arm strength within 72 hours of their symptoms arising to spot evidence of facial asymmetry, arm weakness, and speech abnormalities, which are frequently linked to strokes.

The researchers used Artificial Intelligence techniques to examine 68 facial landmark sites to examine facial asymmetry. Arm weakness was evaluated using a smartphone's built-in 3D accelerometer, gyroscope, and magnetometer. By contrasting normal versus slurred speech patterns, speech modifications were evaluated using mel-frequency cepstral coefficients, which translate sound waves into pictures. The app was then assessed using neurologists' reports and data from brain scans, and it was discovered to reliably identify strokes in almost all instances.

This discovery has great promise for the early detection of stroke symptoms, enabling immediate medical intervention and enhancing survival and recovery prospects. A vascular and interventional neurologist from UCLA's David Geffen School of Medicine, Dr. Radoslav Raychev, showed enthusiasm for the app and using machine learning to treat stroke. This software is intended to revolutionize stroke care and enhance the lives of stroke sufferers.