Fuzzification of the Analysis of Heart Rate Variability Using ECG in Time, Frequency and Statistical Domains

The paper deals with the analysis of heart rate variability in three domains, the time domain, frequency domain and the statistical domain. Different parameters are used in each of the domains and they are given different weights based on their accuracy of detecting arrhythmias. The series of RR intervals of the ECG is used for each of the analysis techniques. The time domain parameters were the standard deviation, average standard deviation of defined segments of the RR intervals, the number of consecutive beats that differ by more than a present value and the RMS value of the standard deviation. The frequency domain analysis included four power spectrums including the FFT, Lomb Periodogram, Burg spectrum and the Yule spectrum. The statistical domain parameters included calculating the Sample entropy, Information Based Similarity Index, Modified Karhunen Loeve Transform Coefficients and viewing Poincare plots. Some of the methods are better analyzing tools as compared to others therefore they are given more weight-age in determining the nature of an RR-interval series. Based on the different weights assigned to the different methods, depending on their credibility, a score is calculated for each of the series of RR-intervals. RR-intervals that get high scores around a pre defined value are considered normal according to the Fuzzification laws.