Intellectual Property India Publishes Patent Application for Privacy-Preserving Federated Healthcare Data Analytics System For Secure Distributed Model Training

MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641118553 A) filed by Ms. Priyanka Chaumwal; A. Lakshmi Bhargav; Dr. K Shravani; Dr. Kamepalli S. L. Prasanna; Dr. N. Vimala; Dr. D. Anupama; B Prasanthi; and Dr. B. Lakshmi Bhavani on October 04, 2026, for Privacy-Preserving Federated Healthcare Data Analytics System For Secure Distributed Model Training.

Inventors include Ms. Priyanka Chaumwal; A. Lakshmi Bhargav; Dr. K Shravani; Dr. Kamepalli S. L. Prasanna; Dr. N. Vimala; Dr. D. Anupama; B Prasanthi; and Dr. B. Lakshmi Bhavani.

The application for the patent was published on October 09, 2026, under issue no. 41/2026.

Abstract: [047] A privacy-preserving federated healthcare data analytics system enables secure distributed model training across independent healthcare nodes without transferring patient-level source data outside respective local data boundaries. A federated learning coordinator distributes a version-controlled global model and training policy. Each participating node locally selects eligible healthcare records, trains the model, bounds a generated model update, and applies an adaptive privacy transformation according to cumulative privacy expenditure, update sensitivity, and local cohort characteristics. A provenance token cryptographically binds the protected update to model and policy versions. A trust validation layer verifies integrity, freshness, version compliance, and anomalous-update characteristics, while a secure aggregation layer combines accepted protected updates without exposing unprotected individual contributions to the coordinator. An adaptive governance engine evaluates privacy expenditure, trust, drift, fairness, convergence, and model performance to modify participation, privacy, clipping, weighting, or validation parameters for subsequent rounds. The architecture thereby integrates federated learning, privacy accounting, provenance validation, secure aggregation, poisoning resistance, and version-controlled governance for multi-institutional healthcare analytics. Accompanied Drawing [FIGS. 1-2].

Disclaimer: Curated by HT Syndication.