Analysis Heartbeat Segmentation with Smoothing Filtering Technique

Abstract

Cardiovascular diseases remain a significant cause of mortality globally, necessitating accurate methods for analyzing cardiac signals for clinical applications such as arrhythmia detection and cardiac disease diagnosis. Heartbeat segmentation is a critical step in delineating different phases of the cardiac cycle, essential for understanding cardiac activity. However, traditional segmentation methods encounter challenges due to noise and artifacts in raw signals, which can compromise accuracy. Smoothing filtering techniques have emerged as a solution to enhance signal quality before segmentation. Among these techniques, the Savitzky-Golay (S-G) filter stands out for its ability to preserve signal characteristics effectively. This study systematically explores the integration of smoothing filters with segmentation algorithms to enhance the accuracy of cardiac signal segmentation, particularly in ambulatory monitoring settings with varying levels of noise and artifacts. Utilizing ECG signals from the MIT-BIH Arrhythmia Database, the study investigates the impact of smoothing filtering on segmentation performance across different signal-to-noise ratios (SNRs). The results demonstrate that smoothing filtering significantly improves segmentation accuracy, particularly at lower SNRs, by mitigating noise-induced inaccuracies. These findings underscore the critical role of smoothing filtering in improving the reliability of cardiac signal analysis, ultimately contributing to enhanced patient care and outcomes in cardiovascular medicine.