MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641092885 A) filed by Presidency University on July 30, 2026, for System And Method For Federated Multi-Modal Ai-Based Tumor Severity Prediction And Personalized Prognosis.
Inventors include Dr. M. Renuka Devi; Dr P Sudha; Suma N G; P. Mahalakshmi; and Pushpalatha M.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: ABSTRACT System and Method for Federated Multi-Modal AI-Based Tumor Severity Prediction and Personalized Prognosis System and method for dynamic severity prediction and personalized prognosis of lung and breast tumors are disclosed. The system comprises a data input module (101) configured to receive multimodal patient data including medical imaging, genomic, proteomic, clinical and laboratory data. A preprocessing module (102) performs normalization, denoising, segmentation and standardization. A feature extraction module (103) derives hand-crafted radiomic features and deep features using neural networks. A multimodal fusion module (105) integrates extracted features and generates a Personalized Patient Knowledge Graph. A severity prediction module (106) computes a dynamically updateable Severity Index Score and classifies tumor states from mild to critical using hybrid ensemble models. A prognostic forecasting module (104) predicts survival probability, progression risk and treatment outcomes using survival modelling and reinforcement learning. A clinical dashboard (107) displays visualizations and explainable AI outputs. The system supports edge/cloud deployment with federated learning and AES-256 encrypted privacy-preserving distributed training.
Disclaimer: Curated by HT Syndication.
