MUMBAI, India, Feb. 13 -- Intellectual Property India has published a patent application (202541124647 A) filed by R. Nithin Kumar; Dr. A. Umamageswari; J. A. Jevin; Ms. R. Thabathi; Ms. K. Deepa; Sikkandhar Batcha J; Pratheeba R S; Renuga R; Krishnavaratha K; and Lakshmi S, Chennai, Tamil Nadu, on Dec. 10, 2025, for 'system and method for deep learning-based analysis of medical images for diagnostic enhancement.'

Inventor(s) include R. Nithin Kumar; Dr. A. Umamageswari; J. A. Jevin; Ms. R. Thabathi; Ms. K. Deepa; Sikkandhar Batcha J; Pratheeba R S; Renuga R; Krishnavaratha K; and Lakshmi S.

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 modular system architecture for deep learning-based medical imaging analysis, designed to enhance diagnostic accuracy and efficiency across diverse radiological modalities. The system integrates a multi-stage pipeline comprising data acquisition, preprocessing, neural network-based feature extraction, and post-processing for clinical decision support. Medical images sourced from PACS or standalone devices are standardized through normalization, denoising, contrast enhancement, and anatomical segmentation. A configurable deep learning framework-featuring convolutional neural networks (CNNs), U-Net, and ResNet architectures-extracts hierarchical features for classification, segmentation, and anomaly detection. Transfer learning and hybrid training strategies improve model generalization under limited labeled data conditions. Post-processing modules generate probability maps, heatmaps, and structured diagnostic reports, which are interoperable with electronic health record (EHR) systems via HL7 and FHIR standards. A feedback loop enables clinician annotations to refine model predictions through active learning. The system demonstrates high diagnostic performance in detecting pathologies such as tumors and fractures, while reducing turnaround time and supporting scalable deployment across healthcare environments. This invention offers a clinically viable, interpretable, and adaptive solution for AI-assisted medical diagnostics."

Disclaimer: Curated by HT Syndication.