Discriminating ECG signals using Support Vector Machines

Nowadays, a reliable electrocardiogram (ECG) analysis and classification plays an important role for diagnosis cardiac abnormalities. Clinically, a computer-assisted technique for ECG analysis can reduced the burden of interpreting the ECG signals. Therefore, this paper proposed a ECG classification analysis using Continuous Wavelet Transform (CWT) and a Support Vector Machine (SVM). CWT is apply to remove noise of ECG signal and to extract distinctive features and used as the inputs to the classifier. SVM was employed merely to classify 4 types of beats of ECG signals namely Normal (N), and three abnormal beats; Left Bundle Branch Block (LBBB), Right Bundle Branch Block (RBBB) and Aberrated Atrial Premature (AAPC). Result obtained indicates that the proposed intelligent discriminating system classified ECG signal types with a high accuracy. The analysis and results also show that the proposed approach is efficient, reliable and applicable.