AI system tracks rehabilitation exercises on simple devices
A lightweight artificial intelligence (AI) system designed to assess rehabilitation exercises in real time could make computer-assisted therapy more practical on low-powered devices, according to research in the International Journal of Business Intelligence and Data Mining.
Researchers have developed a compact artificial intelligence system capable of assessing rehabilitation exercises in real-time on simplistic devices. The lightweight AI, named RMPE Tiny, enhances computer-assisted therapy by operating on low-powered devices typically unable to support precise pose estimation. Utilizing pose estimation, a computer vision method that identifies key body points like shoulders, knees, and ankles, RMPE Tiny measures movement elements such as range, symmetry, and coordination.
The challenge faced by such systems is that high-precision pose estimation requires substantial computing power, limiting their applicability to specialized equipment rather than adaptable devices like tablets or basic monitoring systems. To tackle this issue, the researchers modified RMPE Tiny through several strategies aimed at decreasing computational demands.
These modifications include laser triangulation to improve image acquisition and map 3D coordinates onto a 2D image. In tests, RMPE Tiny achieved an overall pose-estimation accuracy of over 96%, with most samples reaching or surpassing 98%. This system could potentially support rehabilitation programs that offer instant feedback outside clinical environments.
The study detailing RMPE Tiny's design and performance was published in the International Journal of Business Intelligence and Data Mining.
Written by urgent.news from Medical Xpress's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
