Hybrid RR Interval and PCA-Based Feature Extraction for ECG Arrhythmia Classification Using Support Vector Machine

In this study, a hybrid framework for electrocardiogram (ECG)-based cardiac arrhythmia classification is proposed. The primary objective is to develop an efficient approach for distinguishing normal and abnormal heartbeats by integrating temporal and morphological features. The ECG signals were obtained from the standard MIT-BIH Arrhythmia Database. To improve signal quality, the recordings were first preprocessed using a Butterworth band-pass filter, followed by R-peak detection for heartbeat segmentation. Subsequently, RR interval-based features were extracted to characterize the temporal behavior of cardiac activity, while Principal Component Analysis (PCA)-based features were employed to capture the morphological characteristics of the ECG waveform. The integration of these complementary feature sets resulted in a comprehensive feature vector for classification. In the classification stage, a Support Vector Machine (SVM) with a Radial Basis Function (RBF) kernel was utilized. The proposed framework was evaluated using four selected records (106, 119, 208, and 212) from the MIT-BIH Arrhythmia Database. Experimental results demonstrate that the proposed method achieves satisfactory performance in cardiac arrhythmia detection. Furthermore, the combination of temporal and morphological features significantly contributes to improving the overall classification performance, indicating the effectiveness and robustness of the proposed framework for automated ECG analysis