A LIGHTWEIGHT METHOD OF MYOCARDIAL INFARCTION DETECTION & LOCALIZATION FROM SINGLE LEAD ECG FEATURES USING MACHINE LEARNING APPROACH

Abstract

The proposed study titled “A Lightweight Method of Myocardial Infarction Detection & Localization from Single Lead ECG Features Using Machine Learning” focuses on developing an efficient and computationally optimized system for early detection and localization of myocardial infarction (MI) using single-lead Electrocardiogram (ECG) signals. Cardiovascular diseases, particularly myocardial infarction, are among the leading causes of mortality worldwide, and early diagnosis plays a critical role in reducing fatality rates. Traditional diagnostic methods often require multi-lead ECG systems and expert interpretation, which may not be feasible in remote or emergency settings. This research aims to overcome these limitations by utilizing a lightweight machine learning-based approach that can operate effectively on minimal hardware resources. The proposed methodology involves extracting key features from single-lead ECG signals, including time-domain, frequency-domain, and morphological characteristics such as QRS complex, ST-segment deviations, and T-wave variations. These features are preprocessed using noise filtering and normalization techniques to ensure signal quality. Machine learning models such as Support Vector Machine (SVM), Random Forest, and Logistic Regression are employed for classification, while lightweight architectures are prioritized to ensure low computational complexity and fast inference. Additionally, feature selection techniques are applied to reduce dimensionality and improve model efficiency without compromising accuracy. Experimental results demonstrate that the proposed lightweight model achieves high accuracy in detecting myocardial infarction and effectively localizes affected regions using single-lead ECG data. The system performs comparably to more complex multi-lead approaches while maintaining lower computational overhead. This makes it suitable for deployment in portable devices, wearable health monitors, and resource-constrained environments. In conclusion, the proposed approach provides a scalable, cost-effective, and efficient solution for early myocardial infarction detection, enabling timely medical intervention and improving patient outcomes.