Comparison of Time Domain Methods for Alignment of RR Signals Acquired by Different Sensor Systems

This paper compares three different methods for signal alignment called dynamic time warping (DTW), modified dynamic time warping (mDTW) and cross-correlation (CC). RR signals from two different sensor systems, a commercial Wellness Wearable System (Smartex, Italy) and custom-made BACQ system with multi-electrode array (Tecnalia, Sebia) are aligned by DTW, mDTW and CC. Several metrics such as correlation coefficient, reliability, root mean square method and Bland-Altman plots are calculated in order to show the performance of these three methods. All measures indicate that mDTW outperforms other methods. In addition, only mDTW successfully trimmed the beginning and the end of longer signal to match the short one.