MUMBAI, India, March 13 -- Intellectual Property India has published a patent application (202541125809 A) filed by Amrita Vishwa Vidyapeetham, Kollam, Kerala, on Dec. 12, 2025, for 'graph signal processing based system for electroencephalogram classification.'
Inventor(s) include Kalathil, Anuraj; and Menon, Vivek.
The application for the patent was published on March 13, under issue no. 11/2026.
According to the abstract released by the Intellectual Property India: "The present disclosure provides a system (100) for electroencephalogram based classification. The system (100) includes an electroencephalogram acquisition unit (102) to acquire a multi-channel electroencephalogram signal (202), a graph signal processing unit (114) to perform graph learning (208) to derive a functional connectivity network using Learning Graph Laplacian with Gaussian Smoothness prior technique, to perform amplitude pooling (210) to derive pooled amplitude representations, and to perform graph signal modeling (212) by defining the pooled amplitude representations at nodes of the functional connectivity network to generate graph signals. The graph signal processing unit (114) performs scalogram generation (214) to generate graph signal scalograms (216). A classification unit (116) classifies the graph signal scalograms (216) for detection of neurodegenerative disorders. Unlike conventional approaches deriving features using classical signal processing that fail to capture structural relationships across brain regions, the present disclosure jointly represents structural connectivity and functional information for classification."
Disclaimer: Curated by HT Syndication.
