MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641076180 A) filed by Sr University on June 19, 2026, for A Hybrid System For Heart Disease Prediction Using Shap-Based Feature Selection And Pso- Optimization Technique.
Inventors include Kora Swetha; Dr. Vishwanath Bijalwan; and G. Divya Jyothi.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: A HYBRID SYSTEM FOR HEART DISEASE PREDICTION USING SHAP-BASED FEATURE SELECTION AND PSO-OPTIMIZATION TECHNIQUE A hybrid predictive system provides accurate and transparent early heart disease detection by combining SHAP-based feature selection with Particle Swarm Optimization (PSO). The system ingests clinical patient datasets and subjects them to data preprocessing including normalization and missing value handling. A SHAP (SHapley Additive exPlanations) processing pipeline extracts localized and global feature importance metrics to identify and isolate the most impactful cardiac attributes, thereby reducing data dimensionality and eliminating redundant features. A localized metaheuristic optimization engine uses PSO to dynamically explore hyperparameter spaces and minimize classification errors for an integrated machine learning classifier, such as Random Forest or XGBoost. Post-classification, the SHAP engine generates visual and numerical explanations of the individual patient risk predictions. The integrated system drastically minimizes computational complexity and maximizes predictive accuracy, establishing an explainable, scalable, and clinician-friendly decision support tool suitable for real-time diagnostic applications.
Disclaimer: Curated by HT Syndication.
