MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641086942 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on July 16, 2026, for Artificial Intelligence-Based Framework For Automated Detection And Severity Assessment Of Heart Valve Blockages From Echocardiographic Data.

Inventors include Mrs. Soujanya Ambala; B. Rohit; B. Rakesh; M. Akhil; and M. Aman.

The application for the patent was published on July 24, 2026, under issue no. 30/2026.

Abstract: ABSTRACT The present invention discloses an innovative Explainable Artificial Intelligence Framework for Heart Valve Blockage Detection (EHVBD) that enables early, accurate, and transparent identification of heart valve blockages through integrated analysis of static echocardiography images and dynamic video sequences. The proposed system uniquely combines advanced segmentation models such as EfficientNet-based UNet for precise delineation of blocked regions, hybrid classification architectures like Xception-ResNet50 for reliable severity assessment (low, medium, high), and 3D temporal models including 3D ResNet-18 for capturing dynamic cardiac motion patterns. By incorporating comprehensive preprocessing pipelines with augmentation strategies and Grad-CAM-based explainability, the framework generates visual heatmaps that highlight clinically relevant regions influencing predictions, thereby bridging the gap between high-performance deep learning and clinical trust. Deployed via a Flask-based web application, the system supports real-time upload of images and videos, delivers multi-task outputs including segmentation masks, severity classifications, and diagnostic recommendations, and achieves competitive performance metrics such as 72.91% classification accuracy and superior segmentation Dice scores. This holistic, multimodal, and interpretable approach represents a significant advancement in automated cardiovascular diagnostics, facilitating timely interventions and improved patient outcomes.

Disclaimer: Curated by HT Syndication.