Multivariate ensemble classification for the prediction of symptoms in patients with Brugada syndrome

Identification of asymptomatic patients at higher risk for suffering cardiac events remains controversial and challenging in Brugada syndrome (BS). In this work, we proposed an ECG-based classifier to predict BS-related symptoms, by merging the most predictive electrophysiological features derived from the ventricular depolarization and repolarization periods, along with autonomic-related markers. The initial feature space included local and dynamic ECG markers, assessed during a physical exercise test performed in 110 BS patients (25 symptomatic). Morphological, temporal and spatial properties quantifying the ECG dynamic response to exercise and recovery were considered. Our model was obtained by proposing a two-stage feature selection process that combined a resampled-based regularization approach with a wrapper model assessment for balancing, simplicity and performance. For the classification step, an ensemble was constructed by several logistic regression base classifiers, whose outputs were fused using a performance-based weighted average. The most relevant predictors corresponded to the repolarization interval, followed by two autonomic markers and two other makers of depolarization dynamics. Our classifier allowed for the identification of novel symptom-related markers from autonomic and dynamic ECG responses during exercise testing, suggesting the need for multifactorial risk stratification approaches in order to predict future cardiac events in asymptomatic BS patients. Graphical abstract Pipeline for feature selection and predictive modeling of symptoms in Brugada syndrome.

Assessing heart disease using a novel magnetocardiography device

The aim of this paper is to present the use of a portable, unshielded magnetocardiograph (MCG) and identify key characteristics of MCG scans that could be used in future studies to identify parameters that are sensitive to cardiac pathology. We recruited 50 patients with confirmed myocardial infarction (MI) within the past 12 weeks and 46 volunteers with no history of cardiac disease. A set of 38 parameters were extracted from MCG features including both signals from the sensor array and from magnetic images obtained from the device and principal component analysis was used to concentrate the information contained in these parameters into uncorrelated predictors. Linear fits of these parameters were then used to examine the ability of MCG to distinguish between sub-groups of patients. In the fist instance, the primary aim of this study was to ensure that MCG has a basic ability to separate a highly polarised patient group (young controls from post infarction patients) and to identify parameters that could be used in future studies to build a formal diagnostic tool kit. Parameters that parameterised left ventricular ejection fraction (LVEF) were identified and an example is presented to show differential low and high ejection fractions.

ECG Beat Representation and Delineation by Means of Variable Projection

OBJECTIVE

The electrocardiogram (ECG) follows a characteristic shape, which has led to the development of several mathematical models for extracting clinically important information. Our main objective is to resolve limitations of previous approaches, that means to simultaneously cope with various noise sources, perform exact beat segmentation, and to retain diagnostically important morphological information.

METHODS

We therefore propose a model that is based on Hermite and sigmoid functions combined with piecewise polynomial interpolation for exact segmentation and low-dimensional representation of individual ECG beat segments. Hermite and sigmoidal functions enable reliable extraction of important ECG waveform information while the piecewise polynomial interpolation captures noisy signal features like the baseline wander (BLW). For that we use variable projection, which allows the separation of linear and nonlinear morphological variations of the according ECG waveforms. The resulting ECG model simultaneously performs BLW cancellation, beat segmentation, and low-dimensional waveform representation.

RESULTS

We demonstrate its BLW denoising and segmentation performance in two experiments, using synthetic and real data. Compared to state-of-the-art algorithms, the experiments showed less diagnostic distortion in case of denoising and a more robust delineation for the P and T wave.

CONCLUSION

This work suggests a novel concept for ECG beat representation, easily adaptable to other biomedical signals with similar shape characteristics, such as blood pressure and evoked potentials.

SIGNIFICANCE

Our method is able to capture linear and nonlinear wave shape changes. Therefore, it provides a novel methodology to understand the origin of morphological variations caused, for instance, by respiration, medication, and abnormalities.

R-Wave Singularity: A New Morphological Approach to the Analysis of Cardiac Electrical Dyssynchrony

R-wave singularity (RWS) measures the intermittence or discontinuousness of R waves. It has been broadly used in QRS (QRS complex of electrocardiogram) detection, electrocardiogram (ECG) beats classification, etc. In this article, we novelly developed RWS to the analysis of QRS morphology as the measurement of ventricular dyssynchrony and tested the hypothesis that RWS could enhance the discrimination between control and acute myocardial infarction (AMI) patients. Holter ECG recordings were obtained from the Telemetric and Holter ECG Warehouse database, among which database Normal was extracted as normal controls (n = 202) and database AMI (n = 93) as typical subjects of autonomic nervous system dysfunction and cardiac electrical dyssynchrony with high risk for cardiac arrhythmias and sudden cardiac death. Experimental results demonstrate that RWS measured by Lipschitz exponent calculated from 5-min Holter recordings was significantly less negative in early AMI and late AMI than that in Normal subjects for overall, elderly, and elderly male groups, which suggested the heterogeneous depolarization of the ventricular myocardium during AMI. Receiver operating characteristic curve analyses show that combined with heart rate variability parameters, Lipschitz exponent provides higher accuracy in distinguishing between the patients with AMI and healthy control subjects for overall, elderly, elderly male, and elderly female groups. In summary, our study demonstrates the significance of using RWS to probe the cardiac electrical dyssynchrony for AMI. Lipschitz exponent may be valuable and complementary for existing cardiac resynchronization therapy and autonomic nervous system assessment.

Dynamic changes in ventricular depolarization during exercise in patients with Brugada syndrome

Brugada syndrome (BS) is a genetic pathological condition associated with a high risk for sudden cardiac death (SCD). Ventricular depolarization disorders have been suggested as a potential electrophysiological mechanism associated with high SCD risk on patients with BS. This paper aims to characterize the dynamic changes of ventricular depolarization observed during physical exercise in symptomatic and asymptomatic BS patients. To this end, cardiac ventricular depolarization features were automatically extracted from 12-lead ECG recordings acquired during standardized exercise stress test in 110 BS patients, of whom 25 were symptomatic. Conventional parameters were evaluated, including QRS duration, R and S wave amplitudes (AR, AS), as well as QRS morphological features, such as up-stroke and down-stroke slopes of the R and S waves (UR, DR and US). The effects of physical exercise and recovery on the dynamics of these markers were assessed in both BS populations. Features showing significantly different dynamics between the studied groups were used alone and in combination with the clinical characteristics of the patients in a logistic regression analysis. Results show larger changes in the second half of the QRS complex through AS and US measured in the right precordial leads for asymptomatic patients, especially during recovery, when the vagal tone is more pronounced. Multivariate analysis involving both types of features resulted in a reduced model of three relevant features (ΔAS in lead V2, Sex and heart rate recovery, HRR), which achieved a suitable discrimination performance between groups; sensitivity = 80% and specificity = 75% (AUC = 83%). However, after controlling the model for possible confounding factors, only one feature (ΔAS) remained meaningful. This adjusted model significantly improved the overall discrimination performance by up to: sensitivity = 84% and specificity = 100% (AUC = 94%). The study highlights the importance of physical exercise test to unmask differentiated behaviors between symptomatic and asymptomatic BS patients through depolarization dynamic analysis. This analysis together with the obtained model may help to identify asymptomatic patients at low or high risk of future cardiac events, but it should be confirmed by further prospective studies.

Utility of a Smartphone Based System (cvrPhone) to Predict Short-term Arrhythmia Susceptibility

Repolarization alternans (RA) has been implicated in the pathogenesis of ventricular arrhythmias and sudden cardiac death. We developed a 12-lead, blue-tooth/Smart-Phone (Android) based electrocardiogram (ECG) acquisition and monitoring system (cvrPhone), and an application to estimate RA, in real-time. In in-vivo swine studies (N = 17), 12-lead ECG signals were recorded at baseline and following coronary artery occlusion. RA was estimated using the Fast Fourier Transform (FFT) method using a custom developed algorithm in JAVA. Underlying ischemia was detected using a custom developed ischemic index. RA from each lead showed a significant (p < 0.05) increase within 1 min of occlusion compared to baseline (n = 29). Following myocardial infarction, spontaneous ventricular tachycardia episodes (n = 4) were preceded by significant (p < 0.05) increase of RA prior to the onset of the tachy-arrhythmias. Similarly, the ischemic index exhibited a significant increase following myocardial infarction (p < 0.05) and preceding a tachy-arrhythmic event. In conclusion, RA can be effectively estimated using surface lead electrocardiograms by analyzing beat-to-beat variability in ECG morphology using a smartphone based platform. cvrPhone can be used to detect myocardial ischemia and arrhythmia susceptibility using a user-friendly, clinically acceptable, mobile platform.

A Review of Automated Methods for Detection of Myocardial Ischemia and Infarction Using Electrocardiogram and Electronic Health Records

There is a growing body of research focusing on automatic detection of ischemia and myocardial infarction (MI) using computer algorithms. In clinical settings, ischemia and MI are diagnosed using electrocardiogram (ECG) recordings as well as medical context including patient symptoms, medical history, and risk factors-information that is often stored in the electronic health records. The ECG signal is inspected to identify changes in the morphology such as ST-segment deviation and T-wave changes. Some of the proposed methods compute similar features automatically while others use nonconventional features such as wavelet coefficients. This review provides an overview of the methods that have been proposed in this area, focusing on their historical evolution, the publicly available datasets that they have used to evaluate their performance, and the details of their algorithms for ECG and EHR analysis. The validation strategies that have been used to evaluate the performance of the proposed methods are also presented. Finally, the paper provides recommendations for future research to address the shortcomings of the currently existing methods and practical considerations to make the proposed technical solutions applicable in clinical practice.

Identification of exercise-induced ischemia using QRS slopes

In this work we studied a computer-aided approach using QRS slopes as unconventional ECG features to identify the exercise-induced ischemia during exercise stress testing and demonstrated that the performance is comparable to the experts' manual analysis using standard criteria involving ST-segment depression. We evaluated the performance of our algorithm using a database including 927 patients undergoing exercise stress tests and simultaneously collecting the ECG recordings and SPECT results. High resolution 12-lead ECG recordings were collected continuously throughout the rest, exercise, and recovery phases. Patients in the database were classified into three categories of moderate/severe ischemia, mild ischemia, and normal according to the differences in sum of the individual segment scores for the rest and stress SPECT images. Philips DXL 16-lead diagnostic algorithm was run on all 10-s segments of 12-lead ECG recordings for each patient to acquire the representative beats, ECG fiducial points from the representative beats, and other ECG parameters. The QRS slopes were extracted for each lead from the averaged representative beats and the leads with highest classification power were selected. We employed linear discriminant analysis and measured the performance using 10-fold cross-validation. Comparable performance of this method to the conventional ST-segment analysis exhibits the classification power of QRS slopes as unconventional ECG parameters contributing to improved identification of exercise-induced ischemia.

Detection of Faulty Measurements in WBANs using Gaussian Mixture Model and Ant Colony

Wireless sensor networks are subject to different types of faults and interferences after their deployment. Abnormal values reported by sensors should be separated from faulty or injected measurements to ensure reliable monitoring operation. The aim of this paper is to propose a lightweight approach for the detection and suppression of faulty measurements in medical wireless sensor networks. The proposed approach is based on the combination of statistical model and machine learning algorithm. The authors begin by collecting physiological data and then they cluster the data collected during the first few minutes using the Gaussian mixture decomposition. They use the resulted labeled data as the input for the Ant Colony algorithm to derive classification rules in the central base station. Afterward, the derived rules are transmitted and installed in each associated sensor to detect abnormal values in distributed manner, and notify anomalies to the base station. Finally, the authors exploit the spatial and temporal correlations between monitored attributes to differentiate between faulty sensor readings and clinical emergency. They evaluate their approach with real and synthetic patient datasets. The experimental results demonstrate that their proposed approach achieves a high rate of detection accuracy for clinical emergency with reduced false alarm rate when compared to robust Mahalanobis distance.

A novel method to capture the onset of dynamic electrocardiographic ischemic changes and its implications to arrhythmia susceptibility

Background

This study investigates the hypothesis that morphologic analysis of intracardiac electrograms provides a sensitive approach to detect acute myocardial infarction or myocardial infarction‐induced arrhythmia susceptibility. Large proportions of irreversible myocardial injury and fatal ventricular tachyarrhythmias occur in the first hour after coronary occlusion; therefore, early detection of acute myocardial infarction may improve clinical outcomes.

Methods and Results

We developed a method that uses the wavelet transform to delineate electrocardiographic signals, and we have devised an index to quantify the ischemia‐induced changes in these signals. We recorded body‐surface and intracardiac electrograms at baseline and following myocardial infarction in 24 swine. Statistically significant ischemia‐induced changes after the initiation of occlusion compared with baseline were detectable within 30 seconds in intracardiac left ventricle (P<0.0016) and right ventricle–coronary sinus (P<0.0011) leads, 60 seconds in coronary sinus leads (P<0.0002), 90 seconds in right ventricle leads (P<0.0020), and 360 seconds in body‐surface electrocardiographic signals (P<0.0022). Intracardiac leads exhibited a higher probability of detecting ischemia‐induced changes than body‐surface leads (P<0.0381), and the right ventricle–coronary sinus configuration provided the highest sensitivity (96%). The 24‐hour ECG recordings showed that the ischemic index is statistically significantly increased compared with baseline in lead I, aVR, and all precordial leads (P<0.0388). Finally, we showed that the ischemic index in intracardiac electrograms is significantly increased preceding ventricular tachyarrhythmic events (P<0.0360).

Conclusions

We present a novel method that is capable of detecting ischemia‐induced changes in intracardiac electrograms as early as 30 seconds following myocardial infarction or as early as 12 minutes preceding tachyarrhythmic events.

Novel technique for ST-T interval characterization in patients with acute myocardial ischemia

BACKGROUND

The novel signal processing techniques have allowed and improved the use of vectorcardiography (VCG) to diagnose and characterize myocardial ischemia. Herein, we studied vectorcardiographic dynamic changes of ventricular repolarization in 80 patients before (control) and during Percutaneous Transluminal Coronary Angioplasty (PTCA).

METHODS

We propose four vectorcardiographic ST-T parameters, i.e., (a) ST Vector Magnitude Area (aSTVM); (b) T-wave Vector Magnitude Area (aTVM); (c) ST-T Vector Magnitude Difference (ST-TVD), and (d) T-wave Vector Magnitude Difference (TVD). For comparison, the conventional ST-Change Vector Magnitude (STCVM) and Spatial Ventricular Gradient (SVG) were also calculated.

RESULTS

Our results indicate that several vectorcardiographic parameters show significant differences (p-value<0.05) before starting and during PTCA. Statistical minute-by-minute PTCA comparison against the control situation showed that ischemic monitoring reached a sensitivity=90.5% and a specificity=92.6% at the 5th minute of the PTCA, when aSTVM and ST-TVD were used as classifiers.

CONCLUSIONS

We conclude that the sensitivity and specificity for acute ischemia monitoring could be increased with the use of only two vectorcardiographic parameters. Hence, the proposed technique based on vectorcardiography could be used in addition to the conventional ST-T analysis for better monitoring of ischemic patients.

Evaluation of ventricular repolarization dispersion during acute myocardial ischemia: spatial and temporal ECG indices

In this work, we studied the evolution of different electrocardiogram (ECG) indices of ventricular repolarization dispersion (VRD) during acute transmural myocardial ischemia in 95 patients undergoing percutaneous coronary intervention (PCI). We studied both temporal indices of VRD (T-VRD), based on the time intervals of the ECG wave, and spatial indices of VRD (S-VRD), based on the eigenvalues of the spatial correlation matrix of the ECG. The T-wave peak-to-end interval I(TPE) index showed statistically significant differences during left anterior descending artery and right coronary artery (RCA) occlusion for almost the complete time course of the PCI procedure with respect to the control recording. Regarding S-VRD indices, we observed statistically significant increases in the ratio of second to the first eigenvalue I(T21), the ratio of the third to the first eigenvalue I(T31) and the T-wave residuum I(TWR) during RCA occlusions. We also found a statistically significant increase in the I(T31) during left circumflex artery occlusions. To evaluate the evolution of VRD indices during acute ischemia, we calculated the relative change parameter R(I) for each index I. Maximal relative changes (R(I)) during acute ischemia were found for the S-VRD indices I(T21), the first eigenvalue I(λ1) and the second eigenvalue I(λ2), with changes 64, 57 and 52 times their baseline range of variation during the control recording, respectively. Also, we found that relative changes with respect to the baseline were higher in patients with T-wave alternans (TWA) than in those without TWA. In conclusion, results suggest that I(TPE) as well as I(T21), I(T31) and I(TWR) are very responsive to dispersion changes induced by ischemia, but with a behavior which very much depends on the occluded artery.

Acute myocardial ischemia monitoring before and during angioplasty by a novel vectorcardiographic parameter set

BACKGROUND

This work evaluates the vectorcardiographic dynamic changes in ischemic patients before and during Percutaneous Transluminal Coronary Angioplasty (PTCA).

METHODS

Four QRS-loop parameters were computed in 51 ischemic and 52 healthy subjects with the objective of assessing the vectorcardiographic differences between both groups: maximum vector magnitude (QRS(mVM)), planar area (QRS(PA)), maximum distance between centroid and loop (QRS(mDCL)) and perimeter (QRS(P)).The conventional ST-change vector magnitude (STC(VM)), QRS-vector difference (QRS(VD)) and spatial ventricular gradient (SVG) were also calculated.

RESULTS

Statistical minute-by-minute PTCA comparison against a healthy population showed that ischemic patients monitoring is greatly enhanced when all the QRS-loop parameters, in combination with the standard STC(VM), QRS(VD) and SVG indexes, are used in the classification. Sensitivity and Specificity, in turn, reached rather high values, 95.4% and 95.2%, respectively.

CONCLUSIONS

These new vectorcardiographic set of complementary QRS-loop parameters, when combined with the classics STC(VM), QRS(VD) and SVG indexes, increase sensitivity and specificity for acute ischemia monitoring.

Classification of emotional states from electrocardiogram signals: a non-linear approach based on Hurst

Background

Identifying the emotional state is helpful in applications involving patients with autism and other intellectual disabilities; computer-based training, human computer interaction etc. Electrocardiogram (ECG) signals, being an activity of the autonomous nervous system (ANS), reflect the underlying true emotional state of a person. However, the performance of various methods developed so far lacks accuracy, and more robust methods need to be developed to identify the emotional pattern associated with ECG signals.

Methods

Emotional ECG data was obtained from sixty participants by inducing the six basic emotional states (happiness, sadness, fear, disgust, surprise and neutral) using audio-visual stimuli. The non-linear feature ‘Hurst’ was computed using Rescaled Range Statistics (RRS) and Finite Variance Scaling (FVS) methods. New Hurst features were proposed by combining the existing RRS and FVS methods with Higher Order Statistics (HOS). The features were then classified using four classifiers – Bayesian Classifier, Regression Tree, K- nearest neighbor and Fuzzy K-nearest neighbor. Seventy percent of the features were used for training and thirty percent for testing the algorithm.

Results

Analysis of Variance (ANOVA) conveyed that Hurst and the proposed features were statistically significant (p < 0.001). Hurst computed using RRS and FVS methods showed similar classification accuracy. The features obtained by combining FVS and HOS performed better with a maximum accuracy of 92.87% and 76.45% for classifying the six emotional states using random and subject independent validation respectively.

Conclusions

The results indicate that the combination of non-linear analysis and HOS tend to capture the finer emotional changes that can be seen in healthy ECG data. This work can be further fine tuned to develop a real time system.

Electrocardiogram derived respiration from QRS slopes

A method for estimation of respiratory rate from electrocardiogram (ECG) signals, based on variations in slopes of QRS complexes, is presented. 12 standard leads, 3 leads from vectorcardiogram (VCG), and 2 additional non-standard leads derived from VCG loops were analysed. A total of 34 slope series were studied, 2 for each analysed lead: slopes between the peak of Q and R waves, and between the peak of R and S waves. Information of QRS slopes series was combined in order to increase the robustness of estimation. Evaluation is performed over a database containing ECG and respiratory signals simultaneously recorded in 17 subjects spontaneously breathing during a tilt table test. Respiratory rate estimation is performed with information of 4 different combinations of QRS slope series. The best results in respiratory rate estimation error terms are 0.72 ± 4.34%(0.46 ± 7.59 mHz). These results outperform those obtained with other known methods, motivating the use of QRS slopes to obtain reliable respiratory rate estimates.

Very low frequency modulation in QRS slopes and its relation with respiration and heart rate variability during hemodialysis

In this work, we study the very low frequency (VLF) modulation (range 0.01-0.03 Hz) in QRS slopes, heart rate variability (HRV) and ECG-derived respiration in hemodialysis patients. First, the relation between QRS slopes and HRV in the VLF band is measured using ordinary coherence. Then, partial coherence is used to measure the former relationship once the effect related to respiration is removed. Ordinary coherence values above a statistical threshold revealed linear relationship between VLF modulation in QRS slopes and HRV in about 10% of analyzed segments, with mean ± SD values of 0.79 ± 0.07 for upward slope and 0.77 ± 0.06 for downward slope. For these segments, partial coherence values drop below the threshold for 64% of the cases for upward slope and 76% for downward slope, suggesting that the origin of the VLF modulation in QRS slopes is mainly driven by respiration or linearly related to it. In the rest of the cases, partial coherence values dropped with respect to ordinary coherence from 0.89 to 0.77 for upward slope and from 0.86 to 0.75 for downward slope, suggesting that other ANS effects non-linearly related to respiration also contribute to the VLF modulation in QRS slopes.

High-frequency QRS analysis in patients with acute myocardial infarction: a preliminary study

BACKGROUND

The 12-lead electrocardiogram (ECG) is a primary tool in the evaluation and risk stratification of patients with suspected acute myocardial infarction (AMI), even though the initial ECG of these patients is often normal or nondiagnostic. Myocardial ischemia induces depolarization changes that can be quantified by analysis of high-frequency QRS (HFQRS) components. We aimed to demonstrate the potential usefulness of HFQRS analysis in diagnosing myocardial ischemia by characterizing the morphological patterns of the HFQRS signals in patients with AMI before and following reperfusion.

METHODS

Five-minute high-resolution ECG was acquired from 30 patients with AMI (age 55 ± 11 years, 26 men) upon their admission to the intensive coronary care unit (ICCU). Serial ECGs were acquired following coronary revascularization and after additional 24 hours (24h). High-frequency morphology index (HFMI), quantifying the extent of ischemic patterns was computed by a custom software, and its values were compared between the serial ECG measurements.

RESULTS

HFMI values were significantly higher on the admission ECG as compared to the post intervention ECG (4.6 ± 2.9% vs 3.4 ± 2.3%, P < 0.05) and to the 24h ECG (4.6 ± 2.9% vs 2.8 ± 2.1%, P < 0.01). In 79% of the patients who were successfully revascularized HFMI value decreased from admission ECG to 24h ECG.

CONCLUSIONS

Analysis of HFQRS morphology in patients with AMI provides information about the existence and severity of myocardial ischemia. HFQRS analysis may aid in risk stratification of patients with suspected myocardial ischemia, complementarily to conventional ECG.

RELATIONSHIP AMONG QRS COMPLEX CHARACTERS IN ELECTROCARDIOGRAM AND ITS APPLICATION TO MYOCARDIAL ISCHEMIA

Electrocardiogram (ECG) is a noninvasive, economic, and convenient detecting tool in myocardial ischemia (MI), and its clinical appearance is mainly exhibited by ST-T complex changes. Recently, QRS complex characters in detecting MI were proposed by an increasing number of researchers. In this paper, various QRS complex characters were extracted in ECG, and their relationship was analyzed systematically. As a result, these characters were divided into two groups, and there was good correlation among them in each group, while the correlation between the groups was poor. Finally, these QRS complex characters were applied to myocardial ischemia, and five characters had significant differences after 59 normal ECG recordings verification, which were: QRS upward and downward slopes, transient heart rate, angle R and angle Q in a triangle QRS. Experimental results showed it was apparent that the trend changes of these five characters when MI events occurred were consistent with their relationship. The conduction velocity of action potentials in ventricular depolarization is slower in MI states than in normal states.

Myocardial ischemia analysis based on electrocardiogram QRS complex

Electrocardiogram (ECG) is an economic, convenient, and non-invasive detecting tool in myocardial ischemia (MI), and its clinical appearance is mainly exhibited by the changes in ST-T complex. Recently, QRS complex characters were proposed to analyze MI by more and more researchers. In this paper, various QRS complex characters were extracted in ECG signals, and their relationship was analyzed systematically. As a result, these characters were divided into two groups, and there existed good relationship among them for each group, while the poor relationship between the groups. Then these QRS complex characters were applied for statistical analysis on MI, and five characters had significant differences after ECG recording verification, which were: QRS upward and downward slopes, transient heart rate, angle R and angle Q. On the other hand, these QRS complex characters were analyzed in frequency domain. Experimental results showed that the frequency features of RR interval series (Heart Rate Variability, HRV), and QRS barycenter sequence had significant differences between MI states and normal states. Moreover, QRS barycenter sequence performed better.