Biomedical engineers at RMIT University, in collaboration with São Paulo State University, have developed an AI-powered smartphone feature to help paramedics quickly screen for strokes. The tool analyzes facial symmetry and muscle movements, key indicators of stroke, using the Facial Action Coding System. Tested on video recordings, it achieved 82% accuracy in seconds.
The research team aims to collaborate with healthcare providers to develop a mobile application and potentially expand its use for other neurological conditions. Professor Dinesh Kumar highlighted the urgency for real-time diagnostic tools, noting that many strokes are missed in emergency settings and by first responders due to various factors.
This innovation aligns with similar advancements, such as Penn State University and Houston Methodist Hospital's machine learning tool for stroke detection, and AI-driven stroke assessment capabilities in brain scans, like NNS-SOT by Nunaps and AICute by Chulalongkorn University. Additionally, wearable devices with atrial fibrillation detection, such as those by Fitbit and Apple, have gained FDA approval, and mobile apps in Asia, like DrGo and RhythmCam, are incorporating similar functionalities.
Click here to read the original news story.