Multiscale Detection in ECG Multifractal Structure

摘要

In this paper, we introduce a novel measure to characterize multifractality of the electrocardiogram (ECG) signal, the mass exponent spectrum curvature. This parameter reflects nonlinear accumulations of discrepancies of singularity strengths from all adjacent points of the curve. Larger values of it denote stronger multifractality. Moreover, we propose the multiscale analysis method to find the most effective indicators for expressing complex dynamic characteristics during cardiac electrical activities. After carrying out drugs experiments on the mice, we find there always exists a certain scale factor range in which the absolute magnitudes of ECG mass exponent spectra curvature are larger than others, and this range is not changed with data length and maximal scale factor. These conclusions might provide the theoretical bases for early diagnoses and clinical applications.