MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641074395 A) filed by Sasi Institute Of Technology & Engineering India.; S. Karthikeyan; M. Parthiban; P. Sivakumar; and P. Krishnamoorthy on June 16, 2026, for Cross-Cognitive Transfer Learning With Lightweight Deep Learning Models For Explainable Brain Tumor Detection.
Inventors include Sasi Institute Of Technology & Engineering India.; S. Karthikeyan; M. Parthiban; P. Sivakumar; and P. Krishnamoorthy.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: The present invention pertains to an explainable brain tumour detection hybrid deep learning framework based on MRI images. The proposed system combines the capabilities of different deep learning models, including convolutional neural networks (CNNs), Vision Transformers (ViTs), and Cross-Cognitive Transfer Learning (CCTL), to achieve high accuracy and interpretability in tumor classification. The framework is based on local spatio feature extraction and global contextual learning, which are both used to enhance the classification accuracy. Visual interpretation techniques like Grad-CAM and LIME are added to explainable AI techniques to produce visual explanations of tumor regions. Experiments show that the method achieves more accurate, more interpretable and more efficient results than the traditional CNN-only and transformer-only methods. The invention could be used in intelligent healthcare applications or medical imaging applications with the aid of AI.
Disclaimer: Curated by HT Syndication.
