MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641085616 A) filed by Tj Institute Of Technology; Dr. P. Senthil Kumari; Angel. R; Danush. S; and Muthu Selvam on July 13, 2026, for Performance Analysis Of Dilated One -To-Many U-Net Model For Medical Image Segmentation.
Inventors include Dr. P. Senthil Kumari; Angel. R; Danush. S; and Muthu Selvam. G.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The present invention relates to an automated medical image segmentation system using the U-Net deep learning architecture. The system accepts medical images such as MRI, CT, and ultrasound scans, preprocesses them, and performs accurate pixel-level segmentation to identify organs, tissues, or abnormal regions. The generated segmentation masks are refined through post-processing techniques and displayed through a user-friendly Streamlit-based web interface. The invention improves segmentation accuracy, reduces manual effort, minimizes processing time, and supports healthcare professionals in disease diagnosis, treatment planning, and clinical decision making. The proposed system provides an efficient, reliable, and scalable solution for modem medical image analysis. Signature of the Applicant / Authorized Patent Agent Medical image segmentation is a fundamental task in computeraided diagnosis, enabling accurate identification of organs, tissues, and lesions from medical images. Traditional segmentation methods often struggle to capture fine details and complex anatomical structures, resulting in reduced accuracy. This paper presents a Dilated One-to-Many U-Net model that enhances the conventional U-Net architecture by integrating dilated convolutional layers and multiple prediction branches.
Disclaimer: Curated by HT Syndication.
