Empirical mode decomposition in analysis of hemodynamic response to static handgrip

Abstract

The empirical mode decomposition (EMD) of biological signals is used to detect reaction to physiological stimuli and to identify global trends in slowly changing variables. We applied EMD to analyze hemodynamic reaction to handgrip in 9 healthy males (aged 21.3 ± 0.3 years) and 10 male patients following coronary artery bypass grafting (aged 55 ± 6 years). Subjects squeezed a dynamometer with 30% of individually determined maximal force for 3 min. The aim of the study was to check whether the application of EMD to the signals could bring any objective quantitative or qualitative measures allowing one to distinguish physiological states of healthy subjects and patients. Hemodynamic data were collected using a battery-powered, ambulatory impedance cardiography device (ReoMonitor) incorporating a single ECG channel. Heart rate (HR), R-R interval (RR), stroke volume (SV), cardiac output (CO), left ventricular ejection time (ET), pre-ejection period (PEP), maximum amplitude of the dz/dt signal (Amp), and basic chest impedance (Z0) were calculated automatically using the software earlier developed for ReoMonitor. The dedicated computer program allows to calculate and display the dynamics of basic and derivative parameters, describing the impact of systolic time intervals on RR intervals, or the relationship between them (e.g. PEP/ET). The EMD procedure was applied to identify the components of each basic hemodynamic parameter and all their derivatives. We observed the most pronounced effect of handgrip in second and third intrinsic mode functions (IMF), which particularly manifested in parameters describing the ratio of systolic time intervals to the length of RR.