Detection Of Ventricular Late Potentials Research Articles

An efficient machine learning based ventricular late potential detection and classification technique for cardiac healthcare

SUMMARYCardiovascular diseases (CVD) are a major cause of death worldwide each year, with myocardial infarction (MI) accounting for the highest percentage of deaths. The rate of mortality can be reduced if electrical homogeneity of the cardiac tissue is detected early. The presence of the ventricular late potential (VLP) at the conclusion of the QRS complex and the start of the ST segment of the electrocardiogram (ECG) signal indicates a violation. Existing research use a wavelet‐based decomposition of the ECG signal for VLP detection, which is a multistage process. The use of empirical mode decomposition (EMD) to automatically distinguish normal ECG signal and signal with the presence of VLP is proposed in this work. Furthermore, utilizing machine learning classifiers, this research aims to enable automated diagnosis of cardiac arrhythmia. Intrinsic mode functions (IMF) are amplitude and frequency modulated waves that result from the decomposition of ECG using the EMD. Each IMF's peak amplitudes, spectral entropy, and mean of the ECG in the EMD are retrieved and used as input features for various machine learning techniques (ML). According to simulation findings, the bagging classifier achieves accuracy and specificity of 98.93% and 98.62%, respectively, while the AdaBoost classifier based on machine learning technique achieves sensitivity of 99.84%. The observed performance is considered to be the best when compared to existing approaches.

Поздние потенциалы желудочков у детей школьного возраста с сахарным диабетом I типа

Purpose — to study the indicators characterizing the ventricular late potentials (VLP) in schoolchildren with type I diabetes mellitus (TIDM). Materials and methods. 120 children aged 6–17 years were examined (56 — primary school age (6–11 years old), 64 children — senior school age (12–17 years old)). Of which at the age of 6–11 years old: 36 — with TIDM main group I, 20 healthy children (control group I). Of children 12–17 years old, 44 children were with TIDM — main group II, 20 healthy children were included in the control group II. In each age main group рatients with TIDM were divided into 2 subgroups depending on the duration of the disease: subgroup A — 39 patients (22 — primary school age, 17 — senior school age) with diabetes duration 1–3 years and subgroup B — 41 patients (14 и 27 retrospectively) with diabetes duration 4 years or more. All children underwent registration of VLP by high$resolution ECG. The following parameters were calculated: duration of the filtered QRS complex (TotORS, ms), the duration of the low$amplitude (up to 40 μV) part of the signal at the end of the QRS complex (LAS40, ms), the rms amplitude of oscillations in the last 40 ms of the ORS complex (RMS40, μV), as well as the ratio TotQRS / RMS40. The obtained results were processed on the basis of Microsoft Office Excel 2007 spreadsheets using the Statistica 7.0 package for Windows. Results. When analyzing the indicators, it was found that children with diabetes in comparison with healthy children had differences that could be interpreted as signs of lifespan (increase in TotORS, LAS40 and TotQRS / RMS40 ratio, decrease in RMS40 values). Such changes with varying degrees of reliability were recorded in all age groups, and were more pronounced in subgroups B, both in children 6–11 years old, and in patients 12–17 years old. This speaks in favor of the aggravation of the disturbance in the electrical stability of the myocardium as the duration of the disease increases. Conclusions. Thus, the data obtained indicate that children with type I diabetes have foci of delayed fragmented activity due to the electrophysiological heterogeneity of the myocardium. The presence of such foci is recognized as an independent prognostic factor for the risk of fatal arrhythmias, which poses the task of the specialists involved in the management of these patients with the early detection of VLP in order to prevent life$threatening conditions. The research was carried out in accordance with the principles of the Helsinki Declaration. The study protocol was approved by the Local ethics committee of all participating institutions. The informed consent of the patient was obtained for conducting the studies. No conflict of interest was declared by the authors. Key words: type I diabetes mellitus, children, ventricular late potentials, cardiac autonomic neuropathy.

Improving ECG signal denoising using wavelet transform for the prediction of malignant arrhythmias

This paper deals with the accuracy of algorithms for the detection of ventricular late potentials (VLP) in an electrocardiographic signal (ECG), being associated with malignant arrhytmias and possible cardiac death. VLP detection is strongly influenced by signal noise. The objective of this paper is to define a denoising algorithm improving the VLPs detection. The method uses wavelet denoising, subband coding, unfortunately introducing heavy linear distortions. Therefore, an equalisation filter has been properly designed, in order to cancel the distortions. The algorithm has been implemented and successfully verified using MATLAB. Then, it has been implemented on Altera's FPGA and then verified on the evaluation board DE1-SoC. On-board processed results and theoretical results are consistent, validating the algorithm. The results show that the algorithm capability to be implemented as programmable hardware. It also could be used for upgrading ECG devices reliability in the field of heart diseases prevention.

Efficient Detection of Ventricular Late Potentials on ECG Signals Based on Wavelet Denoising and SVM Classification

The analysis of cardiac signals is still regarded as attractive by both the academic community and industry because it helps physicians in detecting abnormalities and improving the diagnosis and therapy of diseases. Electrocardiographic signal processing for detecting irregularities related to the occurrence of low-amplitude waveforms inside the cardiac signal has a considerable workload as cardiac signals are heavily contaminated by noise and other artifacts. This paper presents an effective approach for the detection of ventricular late potential occurrences which are considered as markers of sudden cardiac death risk. Three stages characterize the implemented method which performs a beat-to-beat processing of high-resolution electrocardiograms (HR-ECG). Fifteen lead HR-ECG signals are filtered and denoised for the improvement of signal-to-noise ratio. Five features were then extracted and used as inputs of a classifier based on a machine learning approach. For the performance evaluation of the proposed method, a HR-ECG database consisting of real ventricular late potential (VLP)-negative and semi-simulated VLP-positive patterns was used. Experimental results show that the implemented system reaches satisfactory performance in terms of sensitivity, specificity accuracy, and positive predictivity; in fact, the respective values equal to 98.33%, 98.36%, 98.35%, and 98.52% were achieved.

Marginal Component Analysis of ECG Signals for Beat-to-Beat Detection of Ventricular Late Potentials

Heart condition diagnosis based on electrocardiogram signal analysis is the basic method used in prevention of cardiovascular diseases, which are recognized as the leading cause of death globally. To anticipate the occurrence of ventricular arrhythmia, the detection of Ventricular Late Potentials (VLPs) is clinically worthwhile. VLPs are low-amplitude and high-frequency signals appearing at the end part of QRS complexes in the electrocardiogram, which can be considered as a robust feature for arrhythmia risk stratification in patients with cardiac diseases. This paper proposes a beat-to-beat VLP detection method based on the the marginal component analysis and investigates its performance taking into account different ratios between QRS and VLP power. After a denoising phase, performed adopting the singular vector decomposition technique, heartbeats characterized by VLP onsets are identified and extracted taking into account the vector magnitude of each high resolution ECG (HR-ECG) record. To evaluate the proposed method performance, a 15-lead HR-ECG database consisting of real VLP-negative and simulated VLP-positive patterns was used. The achieved results highlight the method validity for VLP detection.

A Method for Ventricular Late Potentials Detection Using Time-Frequency Representation and Wavelet Denoising

This study proposes a method for ventricular late potentials (VLPs) detection using time-frequency representation and wavelet denoising in high-resolution electrocardiography (HRECG). The analysis is performed both with the signal averaged electrocardiography (SAECG) and in real time. A comparison between the temporal and the time-frequency analysis is also reported. In the first analysis the standard parameters QRSd, LAS40, and RMS40 were used; in the second normalized energy in time-frequency domain was calculated. The algorithm was tested adding artificial VLPs to real ECGs.

Bivariable analysis of ventricular late potentials in high resolution ECG records

In this study the bivariable analysis for ventricular late potentials detection in high-resolution electrocardiographic records is proposed. The standard time-domain analysis and the application of the time-frequency technique to high-resolution ECG records are briefly described as well as their corresponding results. In the proposed technique the time-domain parameter, QRSD and the most significant time-frequency index, ENQRS are used like variables. A bivariable index is defined, that combines the previous parameters. The propose technique allows evaluating the risk of ventricular tachycardia in post-myocardial infarct patients. The results show that the used bivariable index allows discriminating between the patient's population with ventricular tachycardia and the subjects of the control group. Also, it was found that the bivariable technique obtains a good valuation as diagnostic test. It is concluded that comparatively, the valuation of the bivariable technique as diagnostic test is superior to that of the time-domain method and the time-frequency technique evaluated individually.

Quantitative evaluation of a wavelet-based method in ventricular late potential detection

Ventricular late potentials (VLPs) are low-amplitude, high-frequency waveforms appearing in the terminal part of the QRS complex in electrocardiogram (ECG) of patients who are susceptible to ventricular tachycardia and sudden cardiac death, after surviving myocardial infarction. Accordingly, VLP detection presents a prominent non-invasive marker for some cardiac diseases clinically. This paper proposes a VLP detection method based on the wavelet transform and investigates its performance. In this method, a modified vector magnitude waveform is formed using discrete wavelet transform for each high-resolution ECG (HRECG) record; then, by applying the continuous wavelet transform to the QRS complex end part in this waveform, a feature vector is extracted from the resultant time-scale plot. This wavelet-based feature vector is processed by principle component analysis to reduce its dimensionality. Finally, a supervised feedforward artificial neural network, trained by a proper set of these feature vectors, is employed as a classifier. To evaluate the proposed method performance, a HRECG database consisting of the real VLP-negative and simulated VLP-positive patterns is used. In a comparative approach, different VLP detection techniques including the conventional time-domain method, developed by Simson, and some methods utilizing distinct diagnostic features are also applied to this database to investigate the capability of the proposed method in VLP analysis more completely. The results show the proposed method, employing the wavelet transform in both pre-processing and feature extraction stages, reveals high evaluation criteria (accuracy, sensitivity, and specificity) and is qualified to detect VLPs.

Comparative analysis on wavelet-based detection of finite duration low-amplitude signals related to ventricular late potentials

Ventricular late potentials (VLPs) are considered as a noninvasive marker of patients with myocardial infarction, who are prone to the development of ventricular tachycardia. This paper investigates the effects of variations in physical properties of myocardial infarcts in terms of their effects on the parametric variations in VLP analysis. A sufficiently large set of signals underlining the behavior of physical parameters was employed to represent the effect of physical size, position, orientation and type of infarct. The approximated signals are variations from real electrocardiography signals by adding potentials representing late potentials based on duration, frequency, amplitude and position. The aim is not to exactly model VLP but rather to generate an approximate set of signals to examine the performance of the standard methods for different possibilities in infarct dynamics. We investigate some of the detection approaches together with their related assumptions, and try to pinpoint the drawbacks and inaccuracies of these methods and also their assumptions. The three widely accepted criteria—QRS duration, root-mean-square and duration of the signal at the end of QRS for VLP detection—were used in the investigation. Results from the application of these parameters to the set of signals are presented. In addition we investigate the physical nature of an infarct and list a number of possible reasons that might be the cause of a low success rate for the detection of additive potentials. To improve the performance of the common methods, two more wavelet transform parameters are added to those of the standard methods. The method derived from this analysis is presented as an alternative means for the detection of late signals named as delayed potentials, a more general class that includes VLP as a subset.

Reduction of electromyographic noise in the signal-averaged electrocardiogram by spectral decomposition

This paper proposes a technique to improve the quality of high-resolution electrocardiogram by weighting the coherent average of beats by a function of the energy of the corrupting myoelectric noise, prior to subsequent detection of ventricular late potentials. The results obtained with 20 patients indicate the method requires fewer beats than conventional nonweighted average to achieve the same noise level.

Diagnosis and risk stratification in patients at risk of sudden cardiac death

Patients after acute myocardial infarction suffer from an increased risk of sudden cardiac arrhythmogenic death within a time period of one to two years following the acute event. Despite improving strategies for immediate revascularization in the early phase of the infarction (thrombolysis, acute coronary intervention), the rate of sudden cardiac death remains unchanged. Several non-invasive techniques for risk stratification are established: clinical risk profile, measurement of left ventricular function (LV-ejection fraction), resting ECG (detection of repolarization disorders: QT-dispersion, QT-variability), an ECG stress test (detection and severity of myocardial ischemia), ambulatory ECG monitoring (qualitative and quantative analysis of ventricular arrhythmias, QT-variability), surface high resolution ECG (detection of ventricular late potentials), measurement of T-wave alternans (alternans ratio), measurement of the activity and balance of the autonomic nervous system (heart rate variability, baroreflex sensitivity). In addition, programmed ventricular stimulation (PVS) serves as an invasive risk stratification technique (detection of an arrhythmogenic substrate). All of these parameters have been shown to be indicative of an increased risk of sudden cardiac death. However, the prognostic power of the non-invasive techniques is limited. In general, the prognostic value of a negative test is reasonably high (90–100% depending on the test used), whereas the prognostic value of a positive test is rather low (4–42% depending on the test used). The combination of several non-invasive tests may significantly improve the positive predictive value above 50%. However, this goes along with a significant decrease of sensitivity below 50%. Therefore, a combination of non-invasive tests with an invasive electrophysiologic study (PVS) seems reasonable. Furthermore, with regard to cost intensive therapies, such as implantable cardioverter defibrillator or antiarrhythmic therapies with sever side effects, an optimal selection for different therapies is mandatory. Thus, a combination of pre-selecting patients by non-invasive variables, to enhance sensitivity of following PVS to more accurately discriminate low risk from high risk patients, is important for clinical practice.

A novel method for beat-to-beat detection of ventricular late potentials.

A novel method for beat-to-beat detection of ventricular late potentials (VLP) from high-resolution electrocardiograms (ECGs) is presented. ECG signals from the X lead are first filtered using a bandpass filter, and then a time-sequence adaptive filter, to improve its signal-to-noise ratio. Eight features are extracted using wavelet transform, from the VLP time-frequency distribution of the filtered ECG signals, and used as inputs of specially designed artificial neural network for VLP recognition. The artificial neural network was trained and tested using clinical data, respectively. The results show that the presented method can detect beat-to-beat-based VLP with sensitivity of 80% and specificity of 77%, and the detection accuracy is 78%.

Adaptive line enhancing plus modified signalaveraging for ventricular late potential detection

Adaptive line enhancing followed by a modified signal average is proposed, for ventricular late potential detection. The acquisition time is reduced five-fold to ~1 min, while maintaining standards in noise reduction for ventricular late potential analysis.

Evaluation of adaptive line enhancement forbeat-to-beat detection of ventricular late potentials

Ventricular late potentials are usually recovered by signal averaging techniques. Some recent studies have commented on the use of adaptive filtering techniques for the detection of these potentials on a beat-to-beat basis. The author reports an evaluation of the technique of adaptive line enhancement in this role. The ensemble of results suggests that the technique is unsuitable for the application.

Analysis of functional changes in ventricular late potentials for risk assessment of ventricular tachycardias after myocardial infarct

Electrophysiological abnormalities during ischaemia and increased heart rate may influence the detection of ventricular late potentials in the surface electrocardiogram. Whether the analysis of functional changes adds information to the risk stratification of patients prone to ventricular tachycardia is unclear. We therefore retrospectively investigated 100 selected patients (25 with documented, sustained ventricular tachycardia (< 230/min) ( = VT group), 25 resuscitated from ventricular fibrillation (VF group) and 50 without ventricular arrhythmias (phi VT/VF group)) in the chronic phase after myocardial infarction. Late potential analysis was performed at rest, during atrial pacing at a rate of 100/min and 120/min (n = 60), during and after occlusion of the coronary artery for coronary angioplasty (PTCA) (n = 70), and immediately after maximum exercise using selective signal averaging. At rest in 72% of patients in the VT group, in 32% of the VF group, and in 6% of the phi VT/VF group late potentials could be found. During atrial pacing in 80% of patients in the VT group, in 72% of the VF group, and in 10% of the patients in the phi VT/VF group and during ischaemia because of occluded coronary artery in 86% of patients in and the VT group, 70% of the VF group, and in 20% of the patients of the phi VT/VF group late potentials were present. Immediately after maximum exercise which let both ischaemia and increased heart rate, late potentials were detectable in 92% of patients in the VT group, 80% of the VF group, and in 14% of patients in the phi VT/VF group. Similar results could be achieved by using the Holter-ECG after exact correction of recorder tape speed variations. 62% of patients with only by ischaemia, increased heart rate or exercise provokable late potentials and all patients with preexistent not by PTCA extinguished late potentials developed recurrent ventricular tachycardias during the one year follow-up period. Patients without late potentials (n = 50) and patients with preexistent by PTCA extinguished late potentials (n = 11) had no recurrent ventricular tachycardias. Cycle length of recurrent and clinical tachycardia in patients with preexistent not by PTCA extinguished late potentials (n = 18) were significantly longer than in patients with only provokable late potentials (n = 21). Analysis of functional changes of ventricular late potentials with exercise or in Holter ECG recordings promises considerable improvement of postinfarction risk stratification especially in patients prone to ventricular fibrillation.

Prognostic value of late potentials in patients with congestive heart failure.

To investigate whether detection of ventricular late potentials could provide prognostic information in patients with congestive heart failure with or without bundle branch block, we prospectively obtained a signal-averaged ECG from 151 patients with congestive heart failure, using specific criteria in 57 patients with bundle branch block. Late potentials were detected in 49 patients (32.5%); their incidence was not significantly different in patients without (31%; 29 patients) or with bundle branch block (35%; 20 patients). Late potentials were present in 25 of 73 patients (34%) with idiopathic dilated cardiomyopathy, in 20 of 57 patients (35%) with ischaemic cardiomyopathy and in four of 21 patients (19%) with hypertensive heart disease (ns). Age, NYHA class, ejection fraction and use of amiodarone were not statistically different among patients with or without late potentials. In contrast, patients with late potentials had more past episodes of sustained ventricular tachycardia (8.2%; four patients) than those without late potentials (1.9%; two patients). Twenty four hour ambulatory ECGs were obtained in 135 patients (89%). Non-sustained ventricular tachycardia was not correlated with the presence of late potentials found in 45 of 88 patients (51%) without late potentials and in 29 of 47 patients (62%) with late potentials (ns). The mean follow-up was 27 +/- 12 months; 51 patients died, 31 from progressive congestive heart failure and 13 suddenly; seven prospectively had sustained ventricular tachycardia. The total mortality rate, the cardiac mortality rate and sudden death risk were not significantly related to the presence of late potentials; their incidence were respectively 35% (36 patients), 32% (33 patients) and 10% (10 patients) in patients without late potentials and 31% (15 patients), 23% (11 patients) and 6% (three patients) in those without late potentials. The incidence of sustained ventricular tachycardia during follow-up was 2% (two patients) in patients without late potentials and 10% (five patients) in those with late potentials. The incidence of sustained ventricular tachycardia experienced by the patients before the study or seen during follow-up was significantly increased in the presence of late potentials: 18% (nine patients) vs 2% (two patients) in the absence of late potentials (P < 0.001). Removal from the study of data from patients with bundle branch block, patients with severe congestive heart failure (NYHA 3 or 4) or patients taking amiodarone did not alter these results. Thus, signal-averaged ECG results only improved risk stratification for sustained ventricular tachycardia in patients with congestive heart failure and failed to identify patients at high risk for sudden death.

Prognosis and risk stratification after myocardial infarction.

In order to assess prognosis after myocardial infarction, various methods have been used: estimation of ejection fraction, Holter monitoring, detection of ventricular late potentials using signal-averaging techniques, programmed ventricular stimulation of the heart to test the inducibility of ventricular tachyarrhythmias, and parameters that assess heart rate variability. Left ventricular dysfunction is the major determinant for prognosis after myocardial infarction, but left ventricular end-systolic volume may be better than overall ejection fraction. The presence of an 'arrhythmogenic substrate' may be detected by recording low-amplitude, fractionated activity or by artificially introducing premature ventricular extrasystoles using programmed ventricular stimulation. The signal-averaged ECG represents an independent tool for risk assessment. In some studies, signal-averaging and programmed ventricular stimulation proved to be more sensitive for the prediction of sustained ventricular tachycardia than for sudden death. The risk of an arrhythmic event in patients with an abnormal signal-averaged ECG or an abnormal result of programmed ventricular stimulation is increased, although the majority of these patients still do not experience a fatal or life-threatening arrhythmia. This high number of false-positive results limits the practical applicability of the signal-averaged ECG but also of other techniques such as long-term ECG recording. Recent interest has focussed on heart rate variability, which is considered an index of sympatho-vagal interaction after acute myocardial infarction. Patients with decreased heart rate variability have a poorer prognosis than those with normal parameters. Patency of the infarct-related artery is a major factor that governs the potentially beneficial effects of thrombolytic therapy. Many, but not all, studies have shown a reduced prevalence of late potentials after successful thrombolysis or in the presence of an open infarct-related artery. Effective thrombolytic therapy may prevent the development of an abnormal electrophysiological milieu after myocardial infarction. Finally, neurological factors such as depression have been shown to be independent risk factors for mortality after acute myocardial infarction. The role of psycho-social factors and their mechanisms of action need further attention since they open the door for behavioural modification. In conclusion, there are several promising new techniques for identification of patients at risk of ventricular tachyarrhythmias. These techniques seem to be superior to more conventional methods such as long-term ECG monitoring as well as exercise testing. Methods which determine the arrhythmogenic substrate might be combined with a parameter of left ventricular ejection fraction because of the paramount importance of left ventricular pump function. However, due to the large proportion of false-positive results obtained with any of these methods, it seems unlikely that they will clearly surpass the others and be sufficiently accurate for risk assessment in the individual patient.

The signal-averaged ECG: time-domain analysis.

During the past decade, the high-resolution electrocardiogram as a non-invasive technique for the detection of ventricular late potentials has developed from an experimental method into a routinely applied non-invasive method for risk stratification of patients after myocardial infarction. Meanwhile, several approaches have been developed for the detection of ventricular late potentials including time-domain analysis, frequency-domain analysis and spectrotemporal mapping. Clinical applications are no longer limited to patients after myocardial infarction, but cover a wider spectrum of different cardiac diseases. This review focuses on some methodological aspects as well as on the results and current clinical applications of the analysis of the signal-averaged ECG in the time domain.

Ambulatory (Holter) electrocardiography signal-averaging: A current perspective

Signal-averaging is an emerging new application of the ambulatory ECG. Whereas real-time signal-averaging of the high-resolution ECG has led to the measurement of abnormal QRS complex parameters and to the detection of ventricular late potentials in patients with a history of ventricular tachycardia or fibrillation at increased risk of an arrhythmic event, similar measurements can also be derived from ambulatory ECG tape recordings. This review describes the technical differences between real-time and ambulatory high-resolution ECG signal-averaging, and the early clinical studies that correlated measured QRS complex parameters and defined the agreement of late potential detection with the two technologies. Although there appears to be a promising cost-effective benefit from ambulatory ECG signal-averaging, limitations imposed by the technical differences must be recognized, and additional investigation is needed to define the appropriate clinical use and criteria for best diagnostic and prognostic value.

Noninvasive Detection of Ventricular Late Potentials: Direct Comparison of 7 Different High-Gain Recording Systems

Noninvasive Detection of Ventricular Late Potentials: Direct Comparison of 7 Different High-Gain Recording Systems