MUMBAI, India, Sept. 7 -- Intellectual Property India has published a patent application (202641104237 A) filed by Vardhaman College Of Engineering on August 29, 2026, for System And Method For Deep Learning-Based Medical Image Analysis.
Inventors include Mr. Manzoor Mohammad; Mr. Anudeep Meda; Mr. Chittireddy Amarendar Reddy; Ms. Koti Tejasvi; Ms. Depalle Sandhya Rani; and Ms. T Ishverya.
The application for the patent was published on September 04, 2026, under issue no. 36/2026.
Abstract: ABSTRACT The present disclosure relates to a system and method for deep learning-based medical image analysis. A medical image study and acquisition metadata are received, and imaging modality and image-quality characteristics are determined. Preprocessing is selected according to the modality and detected quality state, after which anatomy or candidate regions are localized and analyzed using multi-scale deep-learning inference. Detected or segmented regions are quantitatively measured, and region-level uncertainty is determined together with cross-view, spatial or longitudinal consistency when related images are available. A confidence- gated decision engine combines image quality, inference confidence, uncertainty and available consistency information to determine whether a candidate result is technically validated. Low-confidence regions are selectively reprocessed using changed preprocessing, scale, context or model paths, or an image-reacquisition instruction is generated. Only results satisfying the validation criterion are converted into structured findings linking source-image location, measurements and confidence information for clinician review.
Disclaimer: Curated by HT Syndication.
