MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641094519 A) filed by Sr University on August 04, 2026, for Automated Pneumonia Detection In Chest Radiographs Using Convolutional Neural Networks (cnns).
Inventors include Dr. Basant. Kumar Verma; and Dr. Sandip Bhattacharya.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: The present invention concerns a system and method for automated detection of pneumonia from chest radiograph images using Convolutional Neural Networks (CNNs), deep learning techniques and intelligent medical image analysis. The invention solves the problems of conventional diagnosis of pneumonia, such as delayed detection, low diagnostic accuracy, high false-positive and false- negative rates, and relying on manual analysis of chest X-rays. It accepts chest radiograph images from medical imaging systems, medical database, or cloud storage, then normalizes, filters, segments, enhances, and augments the images. CNN-based deep learning models are trained to detect abnormalities associated with pneumonia, and the processed images are analyzed using these models. The system is able to recognize the images into diagnostic classes: pneumonia-positive, pneumonia-negative, bacterial pneumonia, viral pneumonia, and severity-level classification. The invention also provides generation of prediction scores, confidence values, annotated radiographs and explainable AI-based visualization outputs such as heatmaps and Grad-CAM activation maps. The framework enables adaptive learning, real time screening, and scalability from hospitals, telemedicine, mobile healthcare to resource constrained healthcare environments.
Disclaimer: Curated by HT Syndication.
