MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641076773 A) filed by Vellore Institute Of Technology on June 21, 2026, for A System And Method For Longitudinal Multimodal Progression Prediction Using Temporal Graph-Based Learning.
Inventors include Navamani Tm; and Sudharsan S.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: A longitudinal multimodal progression prediction system (100) and method are disclosed. The system (100) comprises a data acquisition module configured to receive clinical data (104), biomarker data (106), imaging data (108), and longitudinal visit information (110). A preprocessing module (112), temporal harmonization module (114), and state engineering module (116) generate temporally aligned progression representations. Modality-specific encoders (120) produce latent multimodal features that are processed through a patient-specific temporal disease graph construction module (122) and a longitudinal disease trajectory modeling module (124). A stage-aware adaptive multimodal fusion module (126) generates fused representations, which are processed by a multi-horizon forecasting module (128) to generate progression forecasts. An uncertainty-aware prediction module (130) estimates predictive uncertainty associated with generated forecasts. The system further supports explainable feature attribution and adaptive progression analysis through temporal graph-based learning and multimodal representation integration.
Disclaimer: Curated by HT Syndication.
