MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202641044277 A) filed by Saveetha Institute Of Medical And Technical Sciences, Chennai, Tamil Nadu, on April 7, for 'machine learning based fall risk estimation device.'
Inventor(s) include S. Joshua Kumaresan; K. Sathya Siva; and Deepak Nallaswamy Veeraiyan.
The application for the patent was published on May 29, under issue no. 22/2026.
According to the abstract released by the Intellectual Property India: "The proposed invention provides a comprehensive, non-invasive solution for the early detection and prevention of accidental falls through the integration of markerless computer vision and predictive machine learning. Unlike traditional h1ll-detection systems that react only after a traumatic event has occurred, this system focuses on the longitudinal analysis of a subject's locomotion to identify "pre-fall" signatures. The invention utilizes a high-fidelity optical sensor-capable of depth-sensing or standard ROB capture-to monitor a subject's natural movement within a domestic or clinical environment. The core technology lies in its ability to extract precise spatio-temporal gait parameters, such as stride variability, cadence, and gait velocity, alongside critical balance metrics like the center of- mass (CoM) trajectory relative to the base of support. These variables are fed into a machine learning classifier trained on diverse datasets of both healthy and high-risk locomotor patterns. By calculating a multi-factorial fall risk score, the device can detect subtle physiological declines, such as increased mediolateral tmnk sway or shuffling tendencies, which are often invisible to the human eye. This predictive capability enables proactive clinical intervention, allowing caregivers or medical professionals to implement physical therapy or environmental modifications well before a fall incident occurs, thereby significantly improving geriatric safety and long-term1 mobility outcomes."
Disclaimer: Curated by HT Syndication.
