MUMBAI, India, March 13 -- Intellectual Property India has published a patent application (202611010116 A) filed by Shree Guru Gobind Singh Tricentenary University, Gurugram, Haryana, on Jan. 30, for 'federated multi-hospital learning system for disease prediction.'
Inventor(s) include Dr. Yogesh Mehta; and Dr. Ashwini Mehta.
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: "A federated multi-hospital learning system for disease prediction, comprises of a training model configured to enable each hospital to train a local model 101 on its respective patient data, a bias equalization layer 102 for each hospital, a dashboard 104, an aggregation server 105, the training model and the layer 102 are embedded in a non-volatile memory to be executed by a corresponding processor of respective hospital, the layer is configured to compute fairness metrics such as equal opportunity and demographic parity by employing a gradient reweighting and fairness correction protocol 106, the layer 102 employs a local calibration module 103 for correcting local demographic imbalance, comorbidity distribution shifts, and hospital-specific risk distortion, the layer 102 generates the fairness-corrected gradient update 107, the dashboard 104 is configured for inter-hospital equity monitoring, and the data from each hospital is aggregated at the server 105 to generate a global-fairness aware model 108."
Disclaimer: Curated by HT Syndication.
