MUMBAI, India, Sept. 7 -- Intellectual Property India has published a patent application (202641104427 A) filed by Dr. Rvr Nri Institute Of Technology Deemed To Be University; Ms. Chaitanya Yarlagadda; Ms. Sreshta Das Peyyala; Mr. Dhanush Sai Seethala; Mr. Anil Chowdary Makkena; Dr. P. Rajendra Kumar; and Dr. J. Naga Surekha on August 31, 2026, for System And Method For Real-Time Evidential Guidance Of Endoscopic Ultrasound Acquisition Using A Concept-Based Explainability Layer.
Inventors include Ms. Chaitanya Yarlagadda; Ms. Sreshta Das Peyyala; Mr. Dhanush Sai Seethala; Mr. Anil Chowdary Makkena; Dr. P. Rajendra Kumar; and Dr. J. Naga Surekha.
The application for the patent was published on September 04, 2026, under issue no. 36/2026.
Abstract: ABSTRACT A system and method for real-time, closed-loop guidance of endoscopic ultrasound image acquisition is described. Raw ultrasound signal data is processed to create an image, from which a shared feature-extraction backbone and task-specific processing heads produce segmentation and classification outputs. An evidential uncertainty estimation module computes a region-level measure of predictive uncertainty from these outputs, which a concept bottleneck module maps to clear acquisition concepts, such as insufficient acoustic coupling or local boundary-confidence issues, instead of a post-hoc visual heatmap. A per-operator calibration module adjusts thresholds for guidance based on an operator's historical performance, and an operator guidance output module provides corrective instructions to the operator during the acquisition procedure, forming a closed feedback loop. The same concept-level output also fills a structured diagnostic report after acquisition. In an example, the system is applied to endoscopic ultrasound examination of the pancreas, addressing the clear operator dependency and diagnostic challenges typical of that procedure.
Disclaimer: Curated by HT Syndication.
