MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202641097386 A) filed by Vellore Institute Of Technology on August 12, 2026, for System And Method For Topology-Guided Analysis Of Three-Dimensional Medical Images For Lung Nodule Classification.
Inventors include Anitha A; and Sandhiya V.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: A computer-implemented method for topology-guided analysis of a three-dimensional medical image for classifying a lung nodule is disclosed. The method comprises receiving a three-dimensional computed tomography (CT) image comprising a lung region and pre-processing the CT image to generate a standardized tumour volume. A plurality of transformed image representations including a raw intensity representation, a fractional-gradient representation, an edge-enhanced representation, and a normalized structural representation are generated from the standardized tumour volume. Persistent homology descriptors are computed, filtered, and aggregated to generate a unified topological feature representation. Volumetric image features are extracted using a three-dimensional convolutional neural network and fused with the unified topological feature representation to generate a fused feature representation. The lung nodule is classified as benign or malignant based on the fused feature representation, and an explainability output is generated by determining voxel-level attribution values using integrated gradients. FIG. 1
Disclaimer: Curated by HT Syndication.
