Advancing ECG Signal Modeling Through a Hybrid Parametric Spline Approach

Abstract

This paper addresses data privacy concerns and the need for a low computational framework in health sciences education. Introducing hybrid parametric splines as a novel approach in ECG signal modeling, the study explores three cases: parametric cubic spline, parametric quartic spline, and a hybrid approach combining both methods with second-order continuity. Accuracy is assessed using power spectrum analysis, root mean square error (RMSE), percent root mean square difference (PRD), Cross correlation, and Dynamic Time Warping (DTW) validated measures. Additionally, statistical analysis, including the Bland-Altman scatter plot, supports the hybrid approach. The hybrid approach achieves a harmonious blend of smoothness, heightened flexibility, and increased accuracy while ensuring computational simplicity. Efficient utilization of fewer data points optimizes storage and processing. Capable of generating diverse ECG signals, it allows flexibility in creating various scenarios. The hybrid approach demonstrates superior accuracy, as evidenced by the obtained RMSE, PRD, Cross correlation, and DTW values of 0.0410, 15.76, 0.98, and 0.49, respectively.Contributing to the advancement of ECG modeling, these findings provide enhanced visualization, analysis, and educational demonstrations in health science education, particularly in cardiovascular physiology. This research offers valuable insights for improving education in cardiovascular health sciences through the application of ECG modeling. By adopting the hybrid approach, educators and researchers can enhance their understanding and teaching of cardiovascular health, ultimately leading to improved education and advancements in the field.