Automatic Detection of Cardiac Arrhythmias using Discrete Wavelet Transform and Support Vector Machine

Cardiac diseases are a leading cause of mortality globally, necessitating early diagnosis for effective treatment and better patient outcomes. Electrocardiography (ECG) is a widely used diagnostic tool that provides critical insights into the heart's bioelectrical activity. Advances in machine learning have enabled the development of automated systems for detecting cardiac diseases from ECG signals, enhancing timely diagnosis and patient care. This study introduces a method for identifying arrhythmias by employing a Support Vector Machine (SVM) classifier, with feature extraction performed on filtered electrocardiogram (ECG) signals using the Discrete Wavelet Transform (DWT). The model was validated with 3672 ECG datasets from GitHub, using 2931 signals for training and 741 for validation and testing. The model achieved high accuracy, with training and testing accuracies of 99.35% and 97.84%, respectively. Confusion matrix analysis showed minimal misclassification for both training and testing datasets. With an SVM accuracy of 99.3% for arrhythmia detection, the proposed system outperforms other models such as Decision Tree (95.2%), Artificial Neural Network (97.3%), and Random Forest (94.6%). This research highlights the effectiveness of combining DWT and SVM for precise early detection of arrhythmias, helping to lower the risk of severe complications like stroke and cardiac arrest.