MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202611068594 A) filed by National Institute Of Technology, Uttarakhand on June 01, 2026, for Blockchain Integrated Federated Learning System For Privacy-Preserving Histopathology Image Analysis.
Inventors include Dr. Abhay Kumar Yadav; and Dr. Maroti Deshmukh.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The present invention relates to a blockchain integrated federated learning system 100 and method for privacy-preserving histopathology image analysis. The system 100 comprises plurality of decentralized participant nodes 102 interconnected through a blockchain network 104, each participant node 102 performs localized training of a global diagnostic model using institution-specific histopathology image datasets without transmitting raw medical image data outside corresponding institutions. A perceptual-consistency consensus validation module 118 generates perceptual quality scores using Structural Similarity Index and Fréchet Inception Distance evaluations against a Consensus Anchor image dataset stored within a consensus anchor storage unit 120. A blockchain smart contract execution layer 122 validates corresponding model updates according to predefined threshold criteria and selectively accepts or rejects submitted updates. Validated model updates are aggregated through a global aggregation engine 126, while a distributed immutable ledger 124 stores federation metadata, validation outcomes, and node reputation information for secure and auditable federated learning operations.
Disclaimer: Curated by HT Syndication.
