MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641089146 A) filed by Sr University; Samreddy Pooja Reddy; K. Deepa; and Shaik Abdul Nabi on July 22, 2026, for A System And Method For Breast Cancer Classification And Segmentation Using A Pipelined Convolution-Swin Transformer With Anarchic- Beluna Whale Optimized Feed Forward Learning Framework.

Inventors include Samreddy Pooja Reddy; K. Deepa; and Shaik Abdul Nabi.

The application for the patent was published on July 31, 2026, under issue no. 31/2026.

Abstract: The present invention relates to a computer-implemented system and method for automated breast cancer diagnosis using an integrated artificial intelligence framework. Mammogram images are initially preprocessed to improve image quality through noise reduction, contrast enhancement, and normalization. A Pipelined Convolution-Swin Transformer architecture subsequently performs accurate segmentation of suspicious breast lesions by combining convolution-based local feature learning with transformer-based global contextual representation. The segmented lesions are further analyzed using a Multi-Level Covariance Feature Extraction module to obtain compact and discriminative feature representations. Hyperparameters of a Feed Forward Neural Network classifier are automatically optimized using an Anarchic-Beluna Whale Optimization (ABWO) algorithm to enhance learning efficiency and reduce computational complexity. The optimized classifier categorizes mammographic images into normal, benign, or malignant classes. An Explainable Artificial Intelligence module further provides visual interpretation maps highlighting the image regions responsible for the diagnostic decision, thereby improving transparency and clinician confidence. The proposed invention offers enhanced segmentation accuracy, improved classification performance, reduced computational overhead, and greater interpretability, making it suitable for deployment in computer-aided diagnosis systems, hospitals, diagnostic imaging centers, cloud healthcare platforms, and telemedicine applications.

Disclaimer: Curated by HT Syndication.