Influence of the skeletal muscle activity on time and frequency domain properties of the body surface ECG during evolving ventricular fibrillation in the pig

SummaryAim of the studyTo evaluate influence of the skeletal muscle activity (SMA) on time and frequency domain properties of ECG during VF.Materials and methodsWe studied the first 9 min of electrically induced VF (N = 7). We recorded Lead II ECG, 247 unipolar epicardial ventricular electrograms (UEGs) and 3 bipolar skeletal electromyograms (EMGs) near the positions of the ECG electrodes (sampling rate, 500 Hz). We reconstructed ECG (RECG) from UEGs using forward-solution transformation matrix. Spectral properties of ECG, RECG, UEGs and MEGs were assessed in the range 2–250 Hz by the median frequency (MF) and the upper limit of frequency range containing 99% of spectral energy (Flim(99)). Scaling exponent of ECG, RECG and EMGs was calculated in the ranges of 1–8 and 5–20 sampling intervals (ScE1–8 and ScE5–20, respectively).ResultsWe observed non-monotonic increases in MF and Flim(99) of the ECG, but not UEGs and RECG, at 1–5 min of VF. Maximum values of MF and Flim(99) in ECG, UEGs and RECG were (in Hz): 32 ± 29 and 166 ± 67; 11 ± 2 and 36 ± 7; 10 ± 2 and 32 ± 6, respectively. The transient increases in the high-frequency content of the ECG were correlated with enhanced activity in EMGs, characterized by an almost uniform spectrum in the range 2–250 Hz (MF = 92 ± 29; Flim(99) = 245 ± 4 Hz). Peak values of ScE1–8 were the highest in EMGs (1.95 ± 0.04), intermediate in the ECG (1.59 ± 0.26), and the lowest in RECG (1.088 ± 0.007).ConclusionSMA significantly contributes to ECG during VF and can bias metrics used for assessment of VF organization.