MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641094839 A) filed by Dhanalakshmi Srinivasan College Of Engineering And Technology on August 05, 2026, for Ai-Based Early Detection And Risk Prediction Of Jaundice Using Clinical And Liver Function Test Data.
Inventors include Hariharan S; Vaseekaran A; and Malathi G.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The present invention relates to an Artificial Intelligence (AI)-based clinical decision support system for the early detection and risk prediction of j aundice by integrating structured clinical patient information with Liver Function Test (LFT) data. The invention employs supervised machine learning algorithms, including Random Forest, XGBoost, Support Vector Machine (SVM), and Logistic Regression, to analyze patient demographics, clinical symptoms, and biochemical liver parameters for accurate and reliable prediction of jaundice risk. The invention further comprises a comprehensive data preprocessing and feature engineering pipeline that performs data cleaning, missing-value imputation, feature encoding, normalization, feature selection, and feature fusion to generate a unified feature representation for predictive analysis. The processed data are evaluated using multiple machine learning models, and the best-performing model is selected based on standard evaluation metrics, including accuracy, precision, recall, F I -score, and ROC-AUC. The proposed system also incorporates a risk assessment module that generates prediction confidence scores and personalized healthcare recommendations. The invention additionally provides an interactive web-based clinical decision support platform that enables healthcare professionals to securely enter patient information, visualize prediction results, generate diagnostic reports, and monitor patient records. The proposed invention significantly improves the efficiency and accuracy of jaundice diagnosis, supports early medical intervention, reduces diagnostic delays, and provides a scalable, cost-effective, and user-friendly solution suitable for hospitals, diagnostic laboratories, primary healthcare centers, and telemedicine applications. The invention further integrates feature engineering, feature fusion, prediction, visualization, report generation, and security validation modules into a unified and intelligent healthcare framework. These modules collectively enhance the accuracy, reliability, and interpretability of the prediction process by optimizing clinical data, identifying significant health indicators, generating comprehensive diagnostic reports, and presenting the results through an interactive dashboard. The integrated architecture enables healthcare professionals to efficiently analyze patient health status and make evidence-based clinical decisions in a timely manner.
Disclaimer: Curated by HT Syndication.
