MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641119225 A) filed by Sr University on October 05, 2026, for An Integrated Progressive Generative Adversarial Network And Transformer-Capsule System For Automated Lesion Detection And Severity Grading.
Inventors include Dakannagari Harith Reddy; and Anurodh Kumar.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: This invention describes a Progressive GAN and Transformer Capsule-Based System for Lesion Detection and Severity Grading. The system assists in medical image analysis by finding lesions and severity in medical images. It tackles the issue of the scarcity of labelled medical images by generating additional medical images of lesions at various granularities and levels of severity. The system comprises an input module, preprocessing module, image generator, severity simulator, feature extractor, lesion detector, severity classifier and performance evaluator modules. The progressive GAN module generates additional images, while the severity simulator module provides gradual change in lesion properties. The generated image and actual image are analyzed using Transformer and Capsule Network techniques, where the Transformer detects the associations among the parts of the image and the Capsule Network retains the spatial and hierarchical details of lesions. These features are then utilized for detection and severity grading of lesions. Accuracy, precision, recall, F1 score, sensitivity, specificity, and grading error are the metrics employed for the evaluation of the proposed invention. The invention uses image generation, severity simulation, feature extraction, detection, and severity grading.
Disclaimer: Curated by HT Syndication.
