ECG scaling properties of cardiac arrhythmias using detrended fluctuation analysis.

We applied detrended fluctuation analysis to characterize at very short time scales during episodes of cardiac arrhythmias the raw electrocardiogram (ECG) waveform, aiming to get a global insight into its dynamical behaviour in patients who experienced sudden death. We found that in 15 recordings involving different types of arrhythmias (taken from PhysioNet's Sudden Cardiac Death Holter Database), the ECG waveform, besides showing a less-random dynamics, becomes more regular during bigeminy, ventricular tachycardia or even atrial fibrillation and ventricular fibrillation. The ECG waveform scaling properties thus suggest that reduced complexity dominates the underlying mechanisms of arrhythmias. Among other explanations, this may result from shorted or restricted (i.e. less diverse) pathways of conduction of the electrical activity within ventricles.