Receiver-Operating-Characteristic Analysis Reveals Superiority of Scale-Dependent Wavelet and Spectral Measures for Assessing Cardiac Dysfunction
Abstract
Receiver-operating-characteristic (ROC) analysis was used to assess the suitability of various heart rate variability (HRV) measures for correctly classifying electrocardiogram records of varying lengths as normal or revealing the presence of heart failure. Scale-dependent HRV measures were found to be substantially superior to scale-independent measures (scaling exponents) for discriminating the two classes of data over a broad range of record lengths. The wavelet-coefficient standard deviation at a scale near 32 heartbeat intervals, and its spectral counterpart near 1/32 cycles per interval, provide reliable results using record lengths just minutes long. A jittered integrate-and-fire model built around a fractal Gaussian-noise kernel provides a realistic, though not perfect, simulation of heartbeat sequences.
Keywords
Cite
@article{arxiv.physics/9901007,
title = {Receiver-Operating-Characteristic Analysis Reveals Superiority of Scale-Dependent Wavelet and Spectral Measures for Assessing Cardiac Dysfunction},
author = {Stefan Thurner and Markus C. Feurstein and Steven B. Lowen and Malvin C. Teich},
journal= {arXiv preprint arXiv:physics/9901007},
year = {2019}
}
Comments
4 pages, ps figures in a single ps file. Published in Physical Review Letters, Volume 81, Number 25, 5688-5691 (1998)