Ultra-short-term heart rate variability analysis on accelerometric signals from mobile phone

Abstract

The feasibility of measuring stress-related parameters by ultra-short variability (USV) indices calculated from the ballistocardiographic signal acquired by the mobile phone accelerometers (m-BCG) positioned on the navel was tested, and its accuracy compared with gold standard ECG-derived indices. The m-BCG was acquired in six healthy volunteers while in supine position, during spontaneous breathing (CTRL) and during 1 minute of mental stress (MS) induced by arithmetic serial subtraction task. Beat occurrence was independently and automatically extracted from both ECG and m-BCG signals, to compute USV parameters in 30 s intervals, during both the CTRL and MS. Linear regression and Bland-Altman analyses between RR series and m-BCG derived beat-to-beat measurements (JJ series) showed very high correlation (r <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> >0.97), no bias, and narrow limits of agreement (±2SD <; ±34 ms) for both CTRL and MS. A significant decrease (p=0.03 Wilcoxon test) in beat duration, SDNN and RMSSD was found in MS compared to CTRL, in both RR and JJ variability series, underlying the ability of m-BCG in capturing the decrease in parasympathetic tone in agreement with the induced stimulus.