MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641101051 A) filed by Sri Eshwar College Of Engineering on August 20, 2026, for Ai Driven Detection System For Aortic Stenosis.
Inventors include Dr. V. Karunakaran; Ms. Kalaivani T; Ajay Vishnu P; and Madhanika K K.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: The AorticAI diagnostic ecosystem addresses the critical shortage of cardiologists in rural India by implementing a multi-modal, patent-pending framework that integrates systemic hemodynamic biomarkers with high-resolution morphological cardiac imaging to achieve a 99.7% diagnostic precision. The core innovation lies in its "Stage -1" predictive capability, which utilizes an XGBoost regressor to analyze biochemical laboratory data specifically NT-proBNP (N-terminal pro-B-type Natriuretic Peptide), Serum Calcium, LDL Cholesterol, and Systolic Blood Pressure to identify valvular risk profiles before symptomatic lesions manifest in the aortic valve. This systemic risk index is then cross-validated by a Vision Engine powered by a Swin Transformer architecture, which utilizes Shifted Windowing to capture global cardiac structures and Slicing Aided Hyper Inference (SAHI) to detect microscopic early-stage calcification deposits and reduced leaflet excursion that are typically lost during standard echocardiographic image downsampling. By employing a Gated Reciprocal Logic, the system cross-references these two data streams to eliminate false positives in low-risk patients and enhance sensitivity for Stage 4 Severe Aortic Stenosis with hemodynamic compromise. Furthermore, the architecture is optimized for edge-computing environments through 16-bit quantization, allowing the entire diagnostic pipeline to run locally on mobile hardware in offline rural cardiac screening settings. This transparent "Black Box" approach is supported by Explainable AI (XAI) tools, including Grad-CAM heatmaps for valve lesion localization and directional SHAP feature attribution for hemodynamic vital sign analysis, providing rural health workers with a comprehensive, cost-effective, and clinically validated diagnostic shield against preventable cardiac mortality from undetected Aortic Stenosis.
Disclaimer: Curated by HT Syndication.
