MUMBAI, India, Nov. 21 -- Intellectual Property India has published a patent application (202521104869 A) filed by Priyanka V. Deshmukh; Priti V. Kale; Narendra M. Kandoi; and Chinmay V. Patil, Amravati, Maharashtra, on Oct. 30, for 'a system and method for early detection of liver tumors using an adaptive attention-based u-transnet architecture and optimized metaheuristic classifier.'
Inventor(s) include Priti V. Kale; Narendra M. Kandoi; Chinmay V. Patil; and Priyanka V. Deshmukh.
The application for the patent was published on Nov. 21, under issue no. 47/2025.
According to the abstract released by the Intellectual Property India: "The present invention relates to a system and method for the early detection of liver tumors using an adaptive attention-based U-TransNet architecture integrated with an optimized metaheuristic classifier. The invention employs a three-phase approach. In the first phase, an anisotropic diffusion filtering technique is applied to enhance image quality while preserving the structural integrity of medical images. The second phase utilizes a hybrid U-TransNet segmentation model that combines the strengths of U-Net and attention-based transformer networks to achieve accurate tumor boundary delineation and segmentation. In the final phase, a support vector machine classifier optimized through a bio-inspired whale optimization algorithm is used for tumor classification. The proposed invention effectively addresses class imbalance and achieves high diagnostic performance, with an Intersection over Union (IoU) of 98% and a Dice score of 99.01%. The integrated framework demonstrates improved accuracy, precision, recall, and F1-score metrics, each approaching 99%, outperforming existing segmentation and classification methods. This invention provides a reliable, efficient, and interpretable system for early liver tumor detection, thereby aiding clinicians in timely diagnosis and treatment planning."
Disclaimer: Curated by HT Syndication.
