Detecting abnormality in heart dynamics from multifractal analysis of ECG signals

Abstract

The characterization of heart dynamics with a view to distinguish abnormal from normal behavior is an interesting topic in clinical sciences. Here we present an analysis of the Electro-cardiogram (ECG) signals from several healthy and unhealthy subjects using the framework of dynamical systems approach to multifractal analysis. Our analysis differs from the conventional nonlinear analysis in that the information contained in the amplitude variations of the signal is being extracted and quantified. The results thus obtained reveal that the attractor underlying the dynamics of the heart has multifractal structure and the variations in the resultant multifractal spectra can clearly separate healthy subjects from unhealthy ones. We use supervised machine learning approach to build a model that predicts the group label of a new subject with very high accuracy on the basis of the multifractal parameters. By comparing the computed indices in the multifractal spectra with that of beat replicated data from the same ECG, we show how each ECG can be checked for variations within itself. The increased variability observed in the measures for the unhealthy cases can be a clinically meaningful index for detecting the abnormal dynamics of the heart.

Multifractal Spectrum Curvature of RR Tachograms of Healthy People and Patients with Congestive Heart Failure, a New Tool to Assess Health Conditions

“…When we calculate parameter r , we see that all the 54 healthy spectra are biased to the right when the subjects are asleep, and when we analyse 24 h-total time series spectra, five of them are slightly biased to the left and nine spectra corresponding to awake subjects are also slightly biased to the left. In the literature it has been reported that the spectra of healthy subjects are wide [ 8 , 10 , 13 , 15 , 16 ]. And in addition, from our results we observe that healthy subjects’ spectra tend to be biased to the right.…”

“…After that the first findings of multifractality in physiological dynamics were reported by Ivanov et al [ 11 , 12 ], multifractal analysis has been extensively used to study time series obtained from physiological systems [ 6 , 13 , 14 , 15 ], and many other kinds of complex systems with emergent properties such as urban systems [ 10 ], fuel mixtures in internal combustion engines [ 16 ], critical fluctuations in magnetic-field driven random systems [ 17 ], the self-organized social dynamics [ 18 ] and the fluctuations of stock market data [ 1 ], to mention a few. For instance, beat-to-beat RR interval time series are inhomogeneous and non-stationary; they fluctuate in an irregular and complex manner, suggesting that different parts of the signal have different scaling properties [ 8 , 11 , 12 , 13 , 19 , 20 ], including scaling differences associated to sleep-wake, sleep stages and even to circadian phases [ 21 , 22 , 23 , 24 , 25 ]. The multifractality of the heartbeat time series allows us to quantify the greater complexity of healthy dynamics compared to pathological conditions [ 8 , 11 , 15 , 20 , 26 , 27 ].…”

Fractal measures and nonlinear dynamics of overcontact binaries

“…Unlike D 2 described above which is an average measure, the f (α) spectrum takes the local contributions of different regions into account as well. The range of scales in the f (α) spectrum is a good measure of the underling dynamical complexity and has been put to considerable use in various fields [37,43,44].Another important nonlinear measure derived from spectral properties is the bicoherence function, which helps to identify quadratic phase coupling between frequencies in a time series [45]. We specifically mention the main peak bicoherence function, b F ( f ) defined in [42], and used to understand the dynamics of Kepler light curves of RRc Lyrae variable stars [42].In this study, we investigate the light curves of 463 overcontact binary stars in the Kepler field of view [1], using techniques of nonlinear dynamics and search for signs of deterministic chaos by computing their nonlinear measures.…”

“…Unlike D 2 described above which is an average measure, the f (α) spectrum takes the local contributions of different regions into account as well. The range of scales in the f (α) spectrum is a good measure of the underling dynamical complexity and has been put to considerable use in various fields [37,43,44].…”

Bimodality and scaling in recurrence networks from ECG data

“…Most of the studies reported in this direction are on heart rate variability (HRV) data to classify cardiac abnormalities [7,8] and the analysis of the full ECG waveforms remain relatively under-explored [9]. Using multifractal measures it (a) g.ambika@iisertirupati.ac.in (corresponding author) is reported that, the variability in the complexity of the cardiac dynamics is less in the case of healthy subjects as compared to patients [10,11]. But,the conventional techniques of nonlinear time series analysis for multifractal measures require long data sets and hence not very useful for short duration clinical ECG data.…”