MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641096642 A) filed by Vellore Institute Of Technology on August 10, 2026, for System And Method For Multimodal Mri Brain Tumor Segmentation Using Deep Neural Networks.
Inventors include Suganthinic; and Uma Priya D.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: A computer-implemented system (100) and method for multimodal magnetic resonance imaging (MRI) brain tumor segmentation are disclosed. The system (100) comprises a preprocessing module configured to receive MRI data including FLAIR, T1-weighted, T1 contrast-enhanced, and T2-weighted image modalities and generate standardized image data through skull stripping, slice selection, resizing, normalization, and label encoding operations. A segmentation framework processes the standardized image data using a modified U-Net architecture incorporating Sobel edge maps, a volumetric V-Net architecture employing three-dimensional convolutional processing, and a dual-stage W-Net architecture comprising sequential U-Net networks. A performance evaluation module compares segmentation outputs generated by the architectures using segmentation metrics. An output generation module produces segmentation masks corresponding to tumor regions, and a visualization module overlays the segmentation masks on MRI images. The disclosed framework integrates preprocessing, segmentation, evaluation, and visualization within a unified deep-learning architecture.
Disclaimer: Curated by HT Syndication.
