MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641107116 A) filed by Koneru Lakshmaiah Education Foundation on September 06, 2026, for Artificial Intelligence System For Thoracic Abnormality Detection, Visual Localisation And Quantitative Clinical Evaluation.
Inventors include Venkata Sanjeev Kuchun; Sivateja Bandaru; Maganti Praneeth Sri Satya Chowdary; and Monoj Sutradhar.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: The present invention provides an artificial intelligence-based system and method for detecting, classifying and visually localising thoracic abnormalities from chest radiographic images. The system comprises a data acquisition and pre-processing module, a convolutional neural network-based classification module, an explainability module employing Gradient-weighted Class Activation Mapping (Grad-CAM), and an evaluation module. The classification module generates a diagnostic classification and an associated confidence value from a pre-processed chest X-ray image. The explainability module determines class-specific feature importance from gradients of the diagnostic output with respect to convolutional feature maps and generates a localisation heat map identifying regions contributing to the classification. The evaluation module quantitatively compares the generated localisation with clinically annotated reference regions using spatial-overlap measures including Intersection-over-Union, thereby providing interpretable and objectively assessable diagnostic outputs.
Disclaimer: Curated by HT Syndication.
