MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202641071358 A) filed by Cvr College Of Engineering on June 09, 2026, for Improved Stacking-Based Ensemble Learning System For Enhanced Medical Image Analysis.
Inventors include G. Malleswari; and Dr. A. Srinivasa Reddy.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: The current invention is associated with an advanced stacking-style ensemble learning methodology. The system integrates different learning models in order to extract complementary features of medical images. The next step is to merge them using decision-level and feature-level fusion. When the adaptive meta-learner modulates the contribution of each base model through the dynamical nature of its contribution to the overall prediction coupled by the certainty of prediction and the importance of the features, the diagnosis becomes more precise, strong, and is generalized to other imaging modalities. Preprocessing modules improve the quality of image by eliminating noise, standardizing the image, and identifying the region of interest. The ensemble output generating module then produces clinically useful results in the form of segmentation maps, likelihood scores, diagnostic labels or severity indicators. It offers reliable and transparent decision support to determine, classify, and assess the disease progression.
Disclaimer: Curated by HT Syndication.
