MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541123798 A) filed by Dr. A. Vinisha; and P. Lakshmi, Hyderabad, Telangana, on Dec. 9, 2025, for 'brain tumor detection and classification using deep learning methodologies.'
Inventor(s) include Dr. A. Vinisha; and P. Lakshmi.
The application for the patent was published on Jan. 2, under issue no. 01/2026.
According to the abstract released by the Intellectual Property India: "The present invention discloses an advanced, automated system and method for the detection and classification of brain tumors utilizing a novel cascaded deep-learning architecture applied to multi-modal Magnetic Resonance Imaging (MRI) data. The system is designed to address critical limitations in current radiological practice, namely diagnostic latency and subjectivity, by providing a rapid, accurate, and interpretable computer-aided diagnosis tool. The core innovation is a two-stage neural network pipeline. The first stage employs a three-dimensional (3D) modified U-Net model that performs precise voxel-wise segmentation to isolate the tumor mass from surrounding healthy brain tissue across multiple MRI sequences (T1, T1c, T2, and FLAIR). This stage outputs a detailed 3D binary mask delineating the tumor boundaries. The second stage consists of a dedicated 3D Convolutional Neural Network (CNN) classifier equipped with an integrated attention mechanism. This classifier receives only the segmented tumor region-of-interest (ROI), ensuring computational efficiency and forcing the model to learn discriminative features solely from pathological tissue. The attention mechanism provides crucial model interpretability by generating visual heatmaps that identify the specific image regions most influential in the classification decision, such as areas of necrosis or enhancement. The system outputs a comprehensive diagnostic report including the classified tumor type (e.g., glioma, meningioma, pituitary tumor), a confidence metric, the tumor localization mask, and the attention heatmap. This invention significantly enhances clinical workflow by reducing analysis time from hours to minutes, minimizing inter-observer variability, and offering transparent, evidence-based diagnostic support to radiologists, thereby facilitating earlier and more reliable treatment planning for patients with brain tumors."
Disclaimer: Curated by HT Syndication.
