MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641094594 A) filed by Sathyabama Institute Of Science And Technology on August 04, 2026, for Deep Learning-Assisted Mathematical Modeling System And Method For Early Disease Prediction Using Multi-Modal Clinical Data.
Inventors include Dr. Priya Shirley Muller; Dr. S. Maheswari; and Dr. R. Parameswari.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The present invention relates to a deep learning-assisted mathematical modeling system and method for early disease prediction using multi-modal clinical data. The invention integrates heterogeneous healthcare information including electronic health records, laboratory reports, physiological signals, wearable sensor measurements, medical images, genomic information, demographic attributes, medication history, and lifestyle data into a unified predictive framework. A preprocessing engine standardizes the acquired data, after which a deep learning-based feature extraction module automatically generates clinically relevant latent representations. An adaptive mathematical disease modeling engine dynamically optimizes weighting coefficients to compute disease probability, while a prediction engine generates disease occurrence, progression, severity, or recurrence estimates. An explainable artificial intelligence module provides interpretable prediction results through feature importance analysis and confidence estimation. A continuous learning framework incrementally updates model parameters using newly acquired patient information without complete retraining, thereby improving long-term prediction performance. The invention provides enhanced prediction accuracy, robustness against heterogeneous and incomplete datasets, improved clinical interpretability, scalable deployment, secure healthcare data processing, and support for simultaneous prediction of multiple diseases within a unified computational framework. Accompanied Drawing [FIGS. 1-2]
Disclaimer: Curated by HT Syndication.
