MUMBAI, India, Feb. 27 -- Intellectual Property India has published a patent application (202641017393 A) filed by Dr Ramya Rani N; Dr A S Narmadha; Ms Malar; Mr A. Anandkumar; Ms Sandra S; Ms Dharani L; Ms Mohanapriya; and Mr. K. Manoj Kumar, Kinathukadavu, Tamil Nadu, on Feb. 17, for 'system and method for automated pneumonia detection using deep residual learning.'
Inventor(s) include Dr Ramya Rani N; Dr A S Narmadha; Ms Malar; Mr A. Anandkumar; Ms Sandra S; Ms Dharani L; Ms Mohanapriya; and Mr. K. Manoj Kumar.
The application for the patent was published on Feb. 27, under issue no. 09/2026.
According to the abstract released by the Intellectual Property India: "The invention provides a computer-aided diagnostic system and method for the automated detection of pneumonia from chest X-ray images using a deep residual neural network architecture. To address the technical challenge of vanishing gradients in deep networks and the visual similarity between pneumonia and other pulmonary conditions, the system utilizes a ResNet-50 model characterized by residual blocks and skip connections. The system comprises a three-stage processing pipeline, a preprocessing stage that implements a Keras Image Data Generator to perform random data augmentations (including rotations, translations, and flips) to enhance model generalization, a feature extraction stage that leverages transfer learning from a model pre-trained on the ImageNet database to identify high-level radiographic featuresand a classification stage that employs a fully connected layer to map extracted features into binary output classes representing "Pneumonia" or "No Findings." The invention further incorporates Early Stopping and ReduceLROnPlateau logic to dynamically optimize the learning rate and prevent overfitting, thereby providing a high-accuracy, resource-efficient diagnostic tool for clinical decision support."
Disclaimer: Curated by HT Syndication.
