Wavelet transform analysis predicts outcome of DC cardioversion for atrial fibrillation patients

doi:10.1016/j.compbiomed.2006.08.003

Computers in Biology and Medicine

Volume 37, Issue 4, April 2007, Pages 517-523

https://doi.org/10.1016/j.compbiomed.2006.08.003 Get rights and content

Abstract

The aim of this study was to examine whether wavelet transform analysis of the electrocardiogram (ECG) can improve the prediction of the maintenance of sinus rhythm in patients with atrial fibrillation (AF) after external DC cardioversion. We examined a variety of wavelet transform-based statistical markers as potential candidates for the prediction of patient status post-cardioversion. Considering a ‘success’ as a patient who remains in normal sinus rhythm for one month post cardioversion and ‘failure’ as a patient who does not, it was shown the proposed non-parametric classification system can achieve 89% specificity at 100% sensitivity using a non-parametric classification method.

Introduction

Atrial fibrillation (AF) is common, with a prevalence of 0.5% in the adult population, rising to 10% or more in those over 75 years [1]. AF causes fatigue, breathlessness and reduced exercise tolerance, and is associated with a 5–6 fold increase in the incidence of stroke. AF is also associated with increased morbidity, lowered quality of life and development and progression of heart failure. The management of this arrhythmia in the UK alone costs 1% of the NHS budget [2]. Restoration of sinus rhythm with DC cardioversion (i.e. transthoracic electrical countershock) improves symptoms, cardiac output, exercise tolerance and reduces the risk of stroke. In this procedure the patient is given a controlled shock during a general anaesthetic. This is initially successful in around 80–100% of patients. However, approximately 40–60% of cardioverted patients revert to AF within three months subsequent to the procedure and around 60–80% within one year [3], [4], [5], [6].

There is therefore a clear clinical need for a tool which can predict the success of cardioversion of AF patients. The development of such a tool has two-fold benefits, these are:

1.
For the patient, the prevention of a patient being exposed to the risk of cardioversion therapy, including the increased likelihood of a thromboembolic event, when the likelihood of long-term benefits is slim.
2.
For the health care provider, the increased cost effectiveness of the treatments prescribed through a reduction of unproductive theatre time and bed usage.

To date, a number of studies have attempted to find markers for the prediction of cardioversion outcome for AF patients with limited success, including: an elevated level of high sensitivity C-reactive protean [4]; genetic programming algorithms [3]; clinical and echocardiographic markers [6] and Fourier based electrocardiogram (ECG) markers pre-filtered using a template matching strategy for QRS cancellation [5]. These, however, have not proved to have the sensitivity required for clinical use. The absence of a commercially available device to perform this kind of analysis underlines the difficulty in providing a robust technology for the task. In this paper we report on very promising results obtained from an outcome prediction tool based on wavelet transform analysis.

Section snippets

Wavelet transform analysis

A number of alternative time–frequency methods are now available for signal analysis. Of these, the wavelet transform has emerged over recent years as the most favoured tool by researchers for analysing problematic signals across a wide variety of areas in science, engineering and medicine [7]. It is especially valuable because of its ability to elucidate simultaneously local spectral and temporal information from a signal in a more flexible way than alternative time–frequency methods by

Patient data

Forty-four patients from the Royal Infirmary and Western general hospitals in Edinburgh undergoing elective cardioversion for AF had their surface ECG (LII) recorded immediately prior $(< 12 h)$ to receiving the treatment. Of these patients, seven were not included in the study due to: failure to attend follow up clinics (p12, p17), a failure in the collection equipment (p23), atrial flutter (p5), ventricular arrhythmia with ICD (p24), co-existent (intermittent) atrial tachycardia (p37) and an

Results

Fig. 3 contains an overview of the cardioversion patient outcomes through time. Of the 30 patients included in the study, six patients (20%) failed to covert into sinus rhythm. At 4 weeks follow up, 13 (43.3%) patients remained in sinus rhythm and 11 (36.7%) patients reverted to AF. It was found that many of the wavelet markers investigated provided good separation between patient sets. Of these, the peak amplitude and wavelet entropy at low temporal scales provided optimal performance. The

Discussion and conclusions

It was shown that using a non-parametric radial basis neural network classification system can lead to the identification of 15 out of the 17 patients for whom cardioversion failed within one month of the procedure, while capturing all 13 of those who would benefit. In the context of the whole study group, 15 out of all 30 patients (50%) who undertook the procedure could be correctly identified as not benefiting from therapy. The use of the technology in practice would allow a significant

Jamie Watson gained his B.Sc. in Physics from Leicester University and an M.Sc. in Advanced Methods in Computer Science from London University. He also has a Ph.D. in wavelet analysis and neural computer applications in engineering. He has worked on a number of signal processing and physical testing projects both in the public sector and as a scientific officer in the private sector. He is a Director of CardioDigital and is co-inventor of the patent applied for technologies developed by the

References (13)

I.A. Paraskevaidis et al.
Prediction of successful cardioversion and maintenance of sinus rhythm in patients with lone atrial fibrillation
Chest
(2005)
J.N. Watson et al.
Practical issues in the evaluation of methods for the prediction of shock outcome success in out-of-hospital cardiac arrest patients
Resuscitation
(2006)
M.J. Domanski
The epidemiology of atrial fibrillation, Coronary Artery Dis.
(1995)
S. Stewart et al.
Cost of an emerging epidemic: an economic analysis of atrial fibrillation in the UK
Heart
(2004)
P. Zohar et al.
Prediction of maintenance of sinus rhythm after DC cardioversion of atrial fibrillation by non-deterministic modelling
Europace
(2005)
E. Watanabe, T. Arakawa, T. Uchiyama, I. Kodama, H. Hishida, High-sensitivity protein is predictive of successful...

There are more references available in the full text version of this article.

Cited by (14)

Wavelet analysis of cardiac optical mapping data
2015, Computers in Biology and Medicine
Citation Excerpt :
Filtering out all the “noise” frequency components with Fourier-based methods can thus induce loss of information and waveform distortion. The wavelet transform is computed by changing the scale of the mother wavelet, shifting the scaled wavelet in time, multiplying by the signal, and integrating over time [37,38]. The multiresolution algorithm of the wavelet transform decomposes a signal into scales with different time and frequency resolution.
Optical mapping technology is an important tool to study cardiac electrophysiology. Transmembrane fluorescence signals from voltage-dependent dyes need to be preprocessed before analysis to improve the signal-to-noise ratio. Fourier analysis, based on spectral properties of stationary signals, cannot directly provide information on the spectrum changes with respect to time. Fourier filtering has the disadvantage of causing degradation of abrupt waveform changes such as those in action potential signals. Wavelet analysis has the ability to offer simultaneous localization in time and frequency domains, suitable for the analysis and reconstruction of irregular, non-stationary signals like the fast action-potential upstroke, and better than conventional filters for denoising.
We applied discrete wavelet transformation for temporal processing of optical mapping signals and wavelet packet analysis approaches to process activation maps from simulated and experimental optical mapping data from canine right atrium. We compared the results obtained with the wavelet approach to a variety of other methods (Fast Fourier Transformation (FFT) with finite or infinite response filtering, and Gaussian filters).Results: Temporal wavelet analysis improved signal-to-noise ratio (SNR) better than FFT filtering for 5–10 dB SNR, and caused less distortion of the action potential waveform over the full range of simulated noise (5–20 dB). Spatial wavelet filtering produced more efficient denoising and/or more accurate conduction velocity estimates than Gaussian filtering. Propagation patterns were also best revealed by wavelet filtering.Conclusions: Wavelet analysis is a promising tool, facilitating accurate action potential characterization, activation map formation, and conduction velocity estimation.
Characterizing atypical patterns of Heart Rate before Paroxysmal Ventricular Tachycardia
2010, Medical Engineering and Physics
Citation Excerpt :
This is an issue of major practical interest to the extent that predictions are more accurate and fast, and that methods used are noninvasive and inexpensive. In this sense, several measures of Heart Rate Variability (HRV) are powerful tools for the analysis of certain types of arrhythmias [5–14]. However, characteristics of changes in HRV preceding the onset of ventricular arrhythmias are unclear when using conventional techniques [15–18].
Results of a comparative analysis between Heart Rate (HR) patterns occurring before the onset of Paroxysmal Ventricular Tachycardia (PVT) and from healthy subjects are shown. Two study groups were made after electrocardiographic dynamical monitoring (Holter) of volunteers. The first group includes 100 h from 27 healthy control subjects, and the second group consists of 88 h, ending with self-terminating episodes of PVT from 55 patients. Patterns are defined as sequences of consecutive RR intervals, while atypical patterns are defined as corresponding to unlikely behavior of Heart Rate in healthy subjects. We investigated spatial and temporal distributions of these patterns in order to find early signs of PVT. We found that they can be grouped on a reduced number of clusters, and the number of atypical patterns increases as we approach the onset of the episode.
A review on sample entropy applications for the non-invasive analysis of atrial fibrillation electrocardiograms
2010, Biomedical Signal Processing and Control
Citation Excerpt :
However, only 50% (11 out of 22) of patients who relapsed to AF during the first 4 weeks following ECV were correctly discerned. With respect to the use of advanced digital signal processing tools, Watson et al. [84] examined a variety of wavelet transform-based statistical markers as potential candidates to predict the post-cardioversion patient status after 1 month following ECV, and 88% (15 out of 17) sensitivity and 100% (13 out of 13) specificity were obtained. Zohar et al. [85] developed a non-deterministic model for predicting sinus rhythm maintenance after ECV, obtaining 71% (15 out of 21) sensitivity and 96% (22 out of 23) specificity when the patients were followed up during the first 3 months after ECV.
The application of non-linear metrics to physiological signals is a valuable tool because “hidden information” related to underlying mechanisms can be obtained. In this respect, approximate entropy (ApEn) is the most popular non-linear regularity index that has been applied to physiological time series. However, ApEn presents some shortcomings, such as bias, relative inconsistency and dependence on the sample length. A modification of ApEn, named sample entropy (SampEn), was introduced to overcome these deficiencies. Recently, in the context of electrocardiography, SampEn has been applied to study non-invasively atrial fibrillation (AF), which is the most common arrhythmia encountered in clinical practice with unknown mechanisms provoking its onset and termination. Useful clinical information, that could help for a better understanding of AF mechanisms, has been obtained through the application of SampEn to electrocardiographic (ECG) recordings. This work reviews its application in the context of non-invasive analysis of AF. During this arrhythmia, atrial and ventricular components can be regarded as unsynchronized activities, whereby, the application of SampEn to the analysis of each component will be described separately. In first place, clinical challenges in which SampEn has been successfully applied to estimate AF organization from the atrial activity pattern are presented. The AF organization study can provide information on the number of active reentries, which can help to improve AF treatment and to take the appropriate decisions on its management. Next, the heart rate variability study via SampEn, to characterize ventricular response and predict AF onset, is described. Through the aforementioned applications it is remarked throughout this review that SampEn can be considered as a very promising and useful tool towards the non-invasive understanding of AF.
Multi-scale entropy evaluates the proarrhythmic condition of persistent atrial fibrillation patients predicting early failure of electrical cardioversion
2020, Entropy
The Illustrated Wavelet Transform Handbook: Introductory Theory and Applications in Science, Engineering, Medicine and Finance, SECOND EDITION
2017, The Illustrated Wavelet Transform Handbook: Introductory Theory and Applications in Science, Engineering, Medicine and Finance, SECOND EDITION
Identifying stable phase coupling associated with cerebral autoregulation using the synchrosqueezed cross-wavelet transform and low oscillation morlet wavelets
2015, Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS

View all citing articles on Scopus

Paul Addison gained his M.Eng. and Ph.D. degrees from the University of Glasgow. He has authored over one hundred technical papers on a variety of engineering topics and two textbooks concerning wavelet transform topics and fractal geometry. Both are published by the Institute of Physics in the UK. He is currently CEO of CardioDigital Ltd. and is co-inventor of a number of patent applied for technologies in the biosignal analysis field. Professor Addison is both a Chartered Engineer (CEng.) and Chartered Physicist (CPhys.), and is a Fellow of the Institute of Physics (UK).

N. Uchaipichat received the B.Eng. (Electrical Engineering) degree in 1997 from Kasetsart University, Thailand and the M.Eng. (Mechatronics) degree in 1999 from the Asian Institute of Technology, Thailand. He completed his Ph.D. degree at Napier University in Edinburgh in 2005 and is currently a lecturer in Department of Electrical Engineering, Thammasat University, Thailand. His research interests lie in medical signal analysis.

Anoop S. Shah is a House Officer Trainee at the Royal Infirmary of Edinburgh. He graduated in 2006 from Edinburgh University with the degree of MbChb with Honours.

Neil Grubb MD, MRCP is a Consultant Cardiologist in the Cardiovascular Unit at the Royal Infirmary of Edinburgh. His clinical interest is cardiac electrophysiology.

View full text

Wavelet transform analysis predicts outcome of DC cardioversion for atrial fibrillation patients

Abstract

Introduction

Section snippets

Wavelet transform analysis

Patient data

Results

Discussion and conclusions

Chest

Resuscitation

The epidemiology of atrial fibrillation, Coronary Artery Dis.

Cost of an emerging epidemic: an economic analysis of atrial fibrillation in the UK

Heart

Prediction of maintenance of sinus rhythm after DC cardioversion of atrial fibrillation by non-deterministic modelling

Europace