Sudden Cardiac Risk Stratification with Electrocardiographic Indices - A Review on Computational Processing, Technology Transfer, and Scientific Evidence

Machine learning based detection of T-wave alternans in real ambulatory conditions

BACKGROUND AND OBJECTIVE

T-wave alternans (TWA) is a fluctuation in the repolarization morphology of the ECG. It is associated with cardiac instability and sudden cardiac death risk. Diverse methods have been proposed for TWA analysis. However, TWA detection in ambulatory settings remains a challenge due to the absence of standardized evaluation metrics and detection thresholds.

METHODS

In this work we use traditional TWA analysis signal processing-based methods for feature extraction, and two machine learning (ML) methods, namely, K-nearest-neighbor (KNN) and random forest (RF), for TWA detection, addressing hyper-parameter tuning and feature selection. The final goal is the detection in ambulatory recordings of short, non-sustained and sparse TWA events.

RESULTS

We train ML methods to detect a wide variety of alternant voltage from 20 to 100 μV, i.e., ranging from non-visible micro-alternans to TWA of higher amplitudes, to recognize a wide range in concordance to risk stratification. In classification, RF outperforms significantly the recall in comparison with the signal processing methods, at the expense of a small lost in precision. Despite ambulatory detection stands for an imbalanced category context, the trained ML systems always outperform signal processing methods.

CONCLUSIONS

We propose a comprehensive integration of multiple variables inspired by TWA signal processing methods to fed learning-based methods. ML models consistently outperform the best signal processing methods, yielding superior recall scores.

Association between body mass index and electrocardiogram indices: A Mendelian randomization study

INTRODUCTION

Cardiovascular diseases (CVDs) remain a global health concern, and body mass index (BMI) is known to be associated with an increased risk of CVD, but the exact mechanisms underlying this relationship remain unclear. This study employs Mendelian randomization (MR) to investigate the causal association between BMI and electrocardiogram (ECG) indices, providing insights into potential pathways linking obesity to CVD.

METHODS

We conducted a comprehensive MR study utilizing large-scale genetic and ECG data from diverse populations. Instrumental variables were selected from genome-wide association studies, ensuring their relevance to BMI. Causal relationships between BMI and ECG indices, including P wave duration, PR interval, QRS duration, and QT interval, were assessed using various MR methods, with inverse variance weighted (IVW) considered as the primary analysis.

RESULTS

Our MR analysis revealed a significant positive causal association between higher BMI and P wave duration (β = 8.078, 95% CI: 5.322 to 10.833, p < 0.001), suggesting a potential mechanism through which higher BMI may contribute to arrhythmogenic risks. However, no significant causal associations were observed between BMI and PR interval, QRS duration, or QT interval (all p > 0.005). In addition, our study also found that there is no horizontal pleiotropy between BMI and P wave duration, PR interval, QRS duration, and QT interval, suggesting that the conclusions of this study are robust.

CONCLUSION

This study supports a causal relationship between elevated BMI and prolonged P wave duration, a marker of increased atrial arrhythmogenic risk. Further investigations are still needed to elucidate the underlying mechanisms.

Machine Learning approach for TWA detection relying on ensemble data design

Background and objective

T-wave alternans (TWA) is a fluctuation of the ST-T complex of the surface electrocardiogram (ECG) on an every-other-beat basis. It has been shown to be clinically helpful for sudden cardiac death stratification, though the lack of a gold standard to benchmark detection methods limits its application and impairs the development of alternative techniques. In this work, a novel approach based on machine learning for TWA detection is proposed. Additionally, a complete experimental setup is presented for TWA detection methods benchmarking.

Methods

The proposed experimental setup is based on the use of open-source databases to enable experiment replication and the use of real ECG signals with added TWA episodes. Also, intra-patient overfitting and class imbalance have been carefully avoided. The Spectral Method (SM), the Modified Moving Average Method (MMA), and the Time Domain Method (TM) are used to obtain input features to the Machine Learning (ML) algorithms, namely, K Nearest Neighbor, Decision Trees, Random Forest, Support Vector Machine and Multi-Layer Perceptron.

Results

There were not found large differences in the performance of the different ML algorithms. Decision Trees showed the best overall performance (accuracy 0.88 ± 0.04 , precision 0.89 ± 0.05 , Recall 0.90 ± 0.05 , F1 score 0.89 ± 0.03 ). Compared to the SM (accuracy 0.79, precision 0.93, Recall 0.64, F1 score 0.76) there was an improvement in every metric except for the precision.

Conclusions

In this work, a realistic database to test the presence of TWA using ML algorithms was assembled. The ML algorithms overall outperformed the SM used as a gold standard. Learning from data to identify alternans elicits a substantial detection growth at the expense of a small increment of the false alarm.

Mitigating Bias in the Measurement of Heart Rate Variability in Physiological Studies of Spinal Manipulation: A Comparison Between Authentic and Sham Manipulation

OBJECTIVE

The purpose of this study was to identify sources and strategies for the mitigation of bias in studies of spinal manipulation and heart rate variability.

METHODS

A small-scale study compared the effects of a single session of sham and authentic cervical manipulation on heart rate variability as measured by power spectrum analysis. The participants were a sample of 31 healthy young students from the Canadian Memorial Chiropractic College, randomized into 2 study arms. The effectiveness of blinding was evaluated, and 2 alternative methods of data analysis were explored to mitigate risk of bias. Following execution of the study, the stages of implementation and data processing were scored against version 2 of the Cochrane risk-of-bias tool for randomized trials for risk of bias.

RESULTS

The risk of bias arising from (1) the randomization process, (2) missing outcome data, and (3) selection of reported results was judged to be low. Risk of bias in (1) deviations from intended interventions (particularly due to the failure of masking) and (2) the measurement of the outcome, for example, through cleaning of the data, were judged to be high.

CONCLUSION

The use of power spectrum analysis of heart rate variability based on 5-minute recordings of echocardiogram pre-and post-intervention contained multiple sources of bias that were challenging to mitigate. Based upon these findings, power spectrum analysis of heart rate variability using these parameters may be ill-suited to the study of physiological effects of spinal manipulative therapy.

Comparability of Heart Rate Turbulence Methodology: 15 Intervals Suffice to Calculate Turbulence Slope – A Methodological Analysis Using PhysioNet Data of 1074 Patients

Heart rate turbulence (HRT) is a characteristic heart rate pattern triggered by a ventricular premature contraction (VPC). It can be used to assess autonomic function and health risk for various conditions, e.g., coronary artery disease or cardiomyopathy. While comparability is essential for scientific analysis, especially for research focusing on clinical application, the methodology of HRT still varies widely in the literature. Particularly, the ECG measurement and parameter calculation of HRT differs, including the calculation of turbulence slope (TS). In this article, we focus on common variations in the number of intervals after the VPC that are used to calculate TS (#TSRR) posing two questions: 1) Does a change in #TSRR introduce noticeable changes in HRT parameter values and classification? and 2) Do larger values of turbulence timing (TT) enabled by a larger #TSRR still represent distinct HRT? We compiled a free-access data set of 1,080 annotated long-term ECGs provided by Physionet. HRT parameter values and risk classes were determined both with #TSRR 15 and 20. A standard local tachogram was created by averaging the tachograms of only the files with the best heart rate variability values. The shape of this standard VPC sequence was compared to all VPC sequences grouped by their TT value using dynamic time warping (DTW) in order to identify HRT shapes. When calculated with different #TSRR, our results show only a little difference between the number of files with enough valid VPC sequences to calculate HRT (<1%) and files with different risk classes (5 and 6% for HRT0-2 and HRTA-C, respectively). In the DTW analysis, the difference between averaged sequences with a specific TT and the standard sequence increased with increasing TT. Our analysis suggests that HRT occurs in the early intervals after the VPC and TS calculated from late intervals reflects common heart rate variability rather than a distinct response to the VPC. Even though the differences in classification are marginal, this can lead to problems in clinical application and scientific research. Therefore, we recommend uniformly using #TSRR 15 in HRT analysis.

A telediagnosis assistance system for multiple-lead electrocardiography

The diffusion of telemedicine opens-up a new perspective for the development of technologies furthered by Biomedical Engineering. In particular, herein we deal with those related to telediagnosis through multiple-lead electrocardiographic signals. This study focuses on the proof-of-concept of an internet-based telemedicine system as a use case that attests to the feasibility for the development, within the university environment, of techniques for remote processing of biomedical signals for adjustable detection of myocardial ischemia episodes. At each signal lead, QRS complexes are detected and delimited with the J-point marking. The same procedure to detect the complex is used to identify the respective T wave, then the area over the ST segment is applied to detect ischemia-related elevations. The entire system is designed on web-based telemedicine services using multiuser, remote access technologies, and database. The measurements for sensitivity and precision had their respective averages calculated at 11.79 and 24.21% for the leads of lower noise. The evaluations regarding the aspects of user friendliness and the usefulness of the application, resulted in 88.57 and 89.28% of broad or total acceptance, respectively. They are robust enough to enable scalability and can be offered by cloud computing, besides enabling the development of new biomedical signal processing techniques within the concept of distance services, using a modular architecture with collaborative bias.

Age-Related Change in Heart Rate Variability at Resting in Thai Professional Athletes

Background: Although regular exercise is beneficially acknowledged in promoting general health status, the cardiovascular risk of sudden cardiac death in highly training athletes has been reported. Unfortunately, the primary key factor in inducing cardiac problems in athletic populations is still questioned. Objectives: In the present study, we aim to determine factors that could affect the relationship between physical activity level and cardiac autonomic function in the Thai population. Methods: Forty participants were recruited and classified into three groups based on their physical activity level and sport profession, including sedentary men, elite handball players, and professional futsal players. In all participants, we measured predicted maximal oxygen consumption, time- and frequency- domain heart rate variability (HRV) variables at rest. Results: Results demonstrated a positive correlation between time-domain HRV and estimated maximal oxygen consumption, but not with frequency-domain HRV, in which athletes aged lower than 25 years old (18 - 25 years old) had a significantly higher in the standard deviation of normal RR intervals (SDNN) and root mean square of the successive differences (RMSSD) than that in age-matched sedentary subjects. Interestingly, athletes aged above 25 years old (25 - 32 years old) had lower SDNN and RMSSD with higher low frequency/high-frequency ratio than those athletes aged under 25. Since all athletes aged above 25 years old were world-class national futsal players, stress due to the competitive level might be another factor influencing cardiac autonomic activity in athletes. Conclusions: In conclusion, the study suggested the importance of age and stress exposure on cardiac autonomic regulation in elite athletes. The observed correlation also reinforced HRV as a potential marker in detecting abnormal changes in cardiac health even in an athletic population.

Spatial-Temporal Signals and Clinical Indices in Electrocardiographic Imaging (II): Electrogram Clustering and T-wave Alternans

During the last years, attention and controversy have been present for the first commercially available equipment being used in Electrocardiographic Imaging (ECGI), a new cardiac diagnostic tool which opens up a new field of diagnostic possibilities. Previous knowledge and criteria of cardiologists using intracardiac Electrograms (EGM) should be revisited from the newly available spatial-temporal potentials, and digital signal processing should be readapted to this new data structure. Aiming to contribute to the usefulness of ECGI recordings in the current knowledge and methods of cardiac electrophysiology, we previously presented two results: First, spatial consistency can be observed even for very basic cardiac signal processing stages (such as baseline wander and low-pass filtering); second, useful bipolar EGMs can be obtained by a digital processing operator searching for the maximum amplitude and including a time delay. In addition, this work aims to demonstrate the functionality of ECGI for cardiac electrophysiology from a twofold view, namely, through the analysis of the EGM waveforms, and by studying the ventricular repolarization properties. The former is scrutinized in terms of the clustering properties of the unipolar an bipolar EGM waveforms, in control and myocardial infarction subjects, and the latter is analyzed using the properties of T-wave alternans (TWA) in control and in Long-QT syndrome (LQTS) example subjects. Clustered regions of the EGMs were spatially consistent and congruent with the presence of infarcted tissue in unipolar EGMs, and bipolar EGMs with adequate signal processing operators hold this consistency and yielded a larger, yet moderate, number of spatial-temporal regions. TWA was not present in control compared with an LQTS subject in terms of the estimated alternans amplitude from the unipolar EGMs, however, higher spatial-temporal variation was present in LQTS torso and epicardium measurements, which was consistent through three different methods of alternans estimation. We conclude that spatial-temporal analysis of EGMs in ECGI will pave the way towards enhanced usefulness in the clinical practice, so that atomic signal processing approach should be conveniently revisited to be able to deal with the great amount of information that ECGI conveys for the clinician.

On the Robustness of Multiscale Indices for Long-Term Monitoring in Cardiac Signals

The identification of patients with increased risk of Sudden Cardiac Death (SCD) has been widely studied during recent decades, and several quantitative measurements have been proposed from the analysis of the electrocardiogram (ECG) stored in 1-day Holter recordings. Indices based on nonlinear dynamics of Heart Rate Variability (HRV) have shown to convey predictive information in terms of factors related with the cardiac regulation by the autonomous nervous system, and among them, multiscale methods aim to provide more complete descriptions than single-scale based measures. However, there is limited knowledge on the suitability of nonlinear measurements to characterize the cardiac dynamics in current long-term monitoring scenarios of several days. Here, we scrutinized the long-term robustness properties of three nonlinear methods for HRV characterization, namely, the Multiscale Entropy (MSE), the Multiscale Time Irreversibility (MTI), and the Multifractal Spectrum (MFS). These indices were selected because all of them have been theoretically designed to take into account the multiple time scales inherent in healthy and pathological cardiac dynamics, and they have been analyzed so far when monitoring up to 24 h of ECG signals, corresponding to about 20 time scales. We analyzed them in 7-day Holter recordings from two data sets, namely, patients with Atrial Fibrillation and with Congestive Heart Failure, by reaching up to 100 time scales. In addition, a new comparison procedure is proposed to statistically compare the poblational multiscale representations in different patient or processing conditions, in terms of the non-parametric estimation of confidence intervals for the averaged median differences. Our results show that variance reduction is actually obtained in the multiscale estimators. The MSE (MTI) exhibited the lowest (largest) bias and variance at large scales, whereas all the methods exhibited a consistent description of the large-scale processes in terms of multiscale index robustness. In all the methods, the used algorithms could turn to give some inconsistency in the multiscale profile, which was checked not to be due to the presence of artifacts, but rather with unclear origin. The reduction in standard error for several-day recordings compared to one-day recordings was more evident in MSE, whereas bias was more patently present in MFS. Our results pave the way of these techniques towards their use, with improved algorithmic implementations and nonparametric statistical tests, in long-term cardiac Holter monitoring scenarios.

An Interoperable System toward Cardiac Risk Stratification from ECG Monitoring

Many indices have been proposed for cardiovascular risk stratification from electrocardiogram signal processing, still with limited use in clinical practice. We created a system integrating the clinical definition of cardiac risk subdomains from ECGs and the use of diverse signal processing techniques. Three subdomains were defined from the joint analysis of the technical and clinical viewpoints. One subdomain was devoted to demographic and clinical data. The other two subdomains were intended to obtain widely defined risk indices from ECG monitoring: a simple-domain (heart rate turbulence (HRT)), and a complex-domain (heart rate variability (HRV)). Data provided by the three subdomains allowed for the generation of alerts with different intensity and nature, as well as for the grouping and scrutinization of patients according to the established processing and risk-thresholding criteria. The implemented system was tested by connecting data from real-world in-hospital electronic health records and ECG monitoring by considering standards for syntactic (HL7 messages) and semantic interoperability (archetypes based on CEN/ISO EN13606 and SNOMED-CT). The system was able to provide risk indices and to generate alerts in the health records to support decision-making. Overall, the system allows for the agile interaction of research and clinical practice in the Holter-ECG-based cardiac risk domain.

Heart rate variability as predictive factor for sudden cardiac death

Sudden cardiac death (SCD) represents about 25% of deaths in clinical cardiology. The identification of risk factors for SCD is the philosopher's stone of cardiology and the identification of non-invasive markers of risk of SCD remains one of the most important goals for the scientific community.The aim of this review is to analyze the state of the art around the heart rate variability (HRV) as a predictor factor for SCD.HRV is probably the most analyzed index in cardiovascular risk stratification technical literature, therefore an important number of models and methods have been developed.Nowadays, low HRV has been shown to be independently predictive of increased mortality in post- myocardial infarction patients, heart failure patients, in contrast with the data of the general population.Contrariwise, the relationship between HRV and SCD has received scarce attention in low-risk cohorts. Furthermore, in general population the attributable risk is modest and the cost/benefit ratio is not always convenient.The HRV evaluation could become an important tool for health status in risks population, even though the use of HRV alone for risk stratification of SCD is limited and further studies are needed.

Hidden Signals-The History and Methods of Heart Rate Variability

The understanding of heart rate variability (HRV) has increased parallel with the development of modern physiology. Discovered probably first in 1847 by Ludwig, clinical applications evolved in the second part of the twentieth century. Today HRV is mostly used in cardiology and research settings. In general, HRV can be measured over shorter (e.g., 5-10 min) or longer (12 or 24 h) periods. Since 1996, most measurements and calculations are made according to the standard of the Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. As the first step, the series of times between successive R-peaks in the ECG are in milliseconds. It is crucial, however, to identify and remove extrasystoles and artifacts according to standard protocols. The series of QRS distances between successive heartbeats can be analyzed with simple or more sophisticated algorithms, beginning with standard deviation (SDNN) or by the square root of the mean of the sum of squares of differences between adjacent normal RR (rMSSD). Short-term HRV is frequently analyzed with the help of a non-parametric fast Fourier transformation quantifying the different frequency bands during the measurement period. In the last decades, various non-linear algorithms have been presented, such as different entropy and fractal measures or wavelet analysis. Although most of them have a strong theoretical foundation, their clinical relevance is still debated.

A multichannel decision-level fusion method for T wave alternans detection

Sudden cardiac death (SCD) is one of the most prominent causes of death among patients with cardiac diseases. Since ventricular arrhythmia is the main cause of SCD and it can be predicted by T wave alternans (TWA), the detection of TWA in the body-surface electrocardiograph (ECG) plays an important role in the prevention of SCD. But due to the multi-source nature of TWA, the nonlinear propagation through thorax, and the effects of the strong noises, the information from different channels is uncertain and competitive with each other. As a result, the single-channel decision is one-sided while the multichannel decision is difficult to reach a consensus on. In this paper, a novel multichannel decision-level fusion method based on the Dezert-Smarandache Theory is proposed to address this issue. Due to the redistribution mechanism for highly competitive information, higher detection accuracy and robustness are achieved. It also shows promise to low-cost instruments and portable applications by reducing demands for the synchronous sampling. Experiments on the real records from the Physikalisch-Technische Bundesanstalt diagnostic ECG database indicate that the performance of the proposed method improves by 12%-20% compared with the one-dimensional decision method based on the periodic component analysis.

Benchmarking of a T-wave alternans detection method based on empirical mode decomposition

BACKGROUND AND OBJECTIVE

T-wave alternans (TWA) is a fluctuation of the ST-T complex occurring on an every-other-beat basis of the surface electrocardiogram (ECG). It has been shown to be an informative risk stratifier for sudden cardiac death, though the lack of gold standard to benchmark detection methods has promoted the use of synthetic signals. This work proposes a novel signal model to study the performance of a TWA detection. Additionally, the methodological validation of a denoising technique based on empirical mode decomposition (EMD), which is used here along with the spectral method, is also tackled.

METHODS

The proposed test bed system is based on the following guidelines: (1) use of open source databases to enable experimental replication; (2) use of real ECG signals and physiological noise; (3) inclusion of randomized TWA episodes. Both sensitivity (Se) and specificity (Sp) are separately analyzed. Also a nonparametric hypothesis test, based on Bootstrap resampling, is used to determine whether the presence of the EMD block actually improves the performance.

RESULTS

The results show an outstanding specificity when the EMD block is used, even in very noisy conditions (0.96 compared to 0.72 for SNR = 8 dB), being always superior than that of the conventional SM alone. Regarding the sensitivity, using the EMD method also outperforms in noisy conditions (0.57 compared to 0.46 for SNR=8 dB), while it decreases in noiseless conditions.

CONCLUSIONS

The proposed test setting designed to analyze the performance guarantees that the actual physiological variability of the cardiac system is reproduced. The use of the EMD-based block in noisy environment enables the identification of most patients with fatal arrhythmias.

Vagus Nerve Stimulation Modulates Complexity of Heart Rate Variability Differently during Sleep and Wakefulness

Progressive loss of heart rate variability (HRV) and complexity are associated with increased risk of mortality in patients with cardiovascular disease and are a candidate marker for patients at risk of sudden cardiac death. HRV is influenced by the cardiac autonomic nervous system (ANS), although it is unclear which arm of the ANS (sympathetic or parasympathetic) needs to be perturbed to increase the complexity of HRV. In this case–control study, we have analyzed the relation between modulation of vagus nerve stimulation (VNS) and changes in complexity of HRV as a function of states of vigilance. We hypothesize that VNS – being a preferential activator of the parasympathetic system – will decrease the heart rate (HR) and increase the complexity of HRV maximum during sleep. The electrocardiogram (EKG) obtained from a 37-year-old, right-handed male with known intractable partial epilepsy and left therapeutic VNS was analyzed during wakefulness and sleep with VNS ON and OFF states. Age-matched control EKG was obtained from five participants (three with intractable epilepsy and two without epilepsy) that had no VNS implant. The study demonstrated the following: (1) VNS increased the complexity of HRV during sleep and decreased it during wakefulness. (2) An increase in parasympathetic tone is associated with increased complexity of HRV even in the presence of decreased HR. These results need to be replicated in a larger cohort before developing patterned stimulation using VNS to stabilize cardiac dysautonomia and prevent fatal arrhythmias.

Holter Monitoring and Loop Recorders: From Research to Clinical Practice

Holter monitors are tools of proven efficacy in diagnosing and monitoring cardiac arrhythmias. Despite the fact their use is widely prescribed by general practitioners, little is known about their evolving role in the management of patients with cryptogenic stroke, paroxysmal atrial fibrillation, unexplained recurrent syncope and risk stratification in implantable cardioverter defibrillator or pacemaker candidates. New Holter monitoring technologies and loop recorders allow prolonged monitoring of heart rhythm for periods from a few days to several months, making it possible to detect infrequent arrhythmias in patients of all ages. This review discusses the advances in this area of arrhythmology and how Holter monitors have improved the clinical management of patients with suspected cardiac rhythm diseases.