MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641096407 A) filed by Velammal Institute Of Technology; Mrs. Joice Ruby J; Ms. Renuga D; Dr. R. Venkadesh; Gunduboina Vamsi; Koncha Venkata Srinivasulu Reddy; Maddi Venkatesh; and Kaveri Sivakishan Reddy on August 10, 2026, for A Computer Vision Approach For Identifying Cardiac Abnormalities In Ecg Image.
Inventors include Mrs. Joice Ruby J; Ms. Renuga D; Dr. R. Venkadesh; Gunduboina Vamsi; Koncha Venkata Srinivasulu Reddy; Maddi Venkatesh; and Kaveri Sivakishan Reddy.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The present invention relates to a computer vision and artificial intelligence-based system for identifying cardiac abnormalities from Electrocardiogram (ECG) images using image processing, machine learning, and deep learning techniques. Cardiovascular diseases are one of the leading causes of death worldwide, and timely diagnosis of cardiac disorders plays a significant role in improving patient survival rates and treatment outcomes. Conventional ECG interpretation methods rely heavily on manual examination by trained cardiologists, which is time-consuming, prone to human error, and often unavailable in rural or resource-constrained healthcare environments. Existing automated ECG diagnostic systems generally depend on raw ECG signal data and specialized hardware equipment, limiting their accessibility and practical implementation in telemedicine and remote healthcare applications. The proposed invention overcomes these limitations by introducing an intelligent image-based ECG analysis system capable of processing ECG images obtained from scanned reports, screenshots, or digital image sources without requiring raw signal inputs. The system initially collects ECG image datasets containing normal and abnormal cardiac conditions. The acquired ECG images are subjected to multiple preprocessing operations including grayscale conversion, Gaussian filtering, noise removal, contrast enhancement, and image resizing to improve image quality and standardize the input data The invention further employs machine learning and deep learning algorithms such as Support Vector Machine (SVM), XGBoost, AdaBoost, and Convolutional Neural Networks (CNN) for automated classification of ECG images into various cardiac categories including normal heartbeat, abnormal heartbeat, myocardial infarction, and history of myocardial infarction. The CNN model automatically learns spatial and waveform-related features from ECG images, thereby improving prediction performance and reducing the dependency on manual feature engineering. The trained classification model is integrated into a Streamlit-based web application that allows users to upload ECG images and obtain instant diagnostic results in real time through a simple and user-friendly interface. The proposed invention provides several advantages including improved diagnostic accuracy, reduced analysis time, real-time prediction capability, low operational cost, and enhanced accessibility for remote healthcare services. Keywords: Computer Vision, Cardiac abnormalities, ECG image, SVM, CNN,XGBoost, AdaBoost.
Disclaimer: Curated by HT Syndication.
