MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202621094602 A) filed by Symbiosis International Deemed University on August 04, 2026, for Ai-Assisted Multi-Modal Anomaly Fusion System And Method For Monitoring Dementia Patients.
Inventors include Dr. Giridhar Urkude; Shambhavii V. Jaiswal; Vladlen Rebello; and Nikhil Halambi.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: ABSTRACT AI-ASSISTED MULTI-MODAL ANOMALY FUSION SYSTEM AND METHOD FOR MONITORING DEMENTIA PATIENTS An artificial- intelligence assisted multi-modal anomaly fusion system (100) and method for monitoring a dementia patient are disclosed. A sensing subsystem (110) having a motion sensor (111), a current sensor (112), a door sensor (113) and a physiological sensor (114) acquires behavioural and physiological data. An edge processing unit (120) filters the data and extracts a feature vector sent over a transport subsystem (130) to a server (140). The server normalises the feature vector against a circadian baseline (142) and applies a first isolation-forest detector (143), a second isolation-forest detector (144) and an autoencoder detector (145), each outputting a score. A Bayesian fusion engine (150) combines the scores with reliability weights and a dynamic-time-warping comparison (152) to generate a risk score (151). When the risk score exceeds a threshold, an alert module (160) dispatches a natural-language alert to a caregiver device (190), reducing false positives. [
Disclaimer: Curated by HT Syndication.
