MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202641084458 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering Technology on July 09, 2026, for Hybrid Machine Learning And Deep Learning Framework For Accurate Heart Disease Prediction And Clinical Decision Support.

Inventors include Soujanya Ambala; G. Abhi Sree; M. Padma Tejaswi; M. Anvita; M. Srija; and Shreya Sridhar.

The application for the patent was published on July 17, 2026, under issue no. 29/2026.

Abstract: ABSTRACT [0027] The present invention introduces an Explainable AI-Driven Hybrid Ensemble System for Early Cardiac Risk Prediction (EAI-HES) that intelligently combines multiple machine learning and deep learning paradigms to deliver highly accurate, transparent, and clinically actionable predictions of heart disease risk. The proposed system processes structured patient data through a comprehensive preprocessing pipeline and employs a novel stacked hybrid architecture wherein Random Forest and Artificial Neural Network base models generate intermediate predictions that are subsequently refined by an XGBoost meta-learner. This layered ensemble approach captures both interpretable decision boundaries and complex non-linear patterns inherent in medical datasets. To overcome the black-box limitations of conventional models, the invention integrates SHAP for global and local feature attribution alongside LIME for patient-specific explanations, enabling physicians to understand the contribution of individual risk factors such as age, cholesterol, blood pressure, and lifestyle indicators. Implemented as a full-stack web application with Flask backend and intuitive frontend, the system classifies patients into low, medium, or high risk categories while providing visual and textual rationales. Extensive preprocessing including imputation, encoding, scaling, and SMOTE balancing, coupled with rigorous evaluation, yields superior performance with XGBoost achieving approximately 95% accuracy. This invention advances cardiovascular diagnostics by merging ensemble robustness, deep representational learning, and explainability, thereby fostering trust and facilitating early intervention in clinical practice.

Disclaimer: Curated by HT Syndication.