MUMBAI, India, Feb. 13 -- Intellectual Property India has published a patent application (202641008124 A) filed by Koneru Lakshmaiah Education Foundation; G. Manu; Sravan Kumar Gunturi; Hari Prasada Raju Kunadharaju; and Ramakrishna Akella, Hyderabad, Telangana, on Jan. 28, for 'system and method for non-invasive detection and classification of colorectal cancer using a hybrid ensemble deep learning framework.'
Inventor(s) include G. Manu; Sravan Kumar Gunturi; Hari Prasada Raju Kunadharaju; and Ramakrishna Akella.
The application for the patent was published on Feb. 13, under issue no. 07/2026.
According to the abstract released by the Intellectual Property India: "The present invention discloses a non-invasive computer-aided diagnosis system for detecting and classifying colorectal cancer using a hybrid ensemble deep learning framework. The system integrates multi-scale convolutional neural networks for local feature extraction, transformer-based architectures for global context modeling, and attention-based multiple instance learning for weakly supervised region selection. An adaptive ensemble learning strategy fuses outputs of individual models to generate a final diagnostic confidence score. Experimental evaluation demonstrates that the proposed system achieves an overall accuracy of approximately 96.4%, sensitivity of 95.8%, specificity of 96.9%, and ROC-AUC of 0.98, outperforming existing state-of-the-art models. The proposed system provides reliable, explainable, and real-time clinical decision support for early colorectal cancer detection and improved patient outcomes."
Disclaimer: Curated by HT Syndication.
