The electrocardiogram as a prognostic tool for predicting major cardiac events

A deep learning-based electrocardiogram risk score for long term cardiovascular death and disease

The electrocardiogram (ECG) is the most frequently performed cardiovascular diagnostic test, but it is unclear how much information resting ECGs contain about long term cardiovascular risk. Here we report that a deep convolutional neural network can accurately predict the long-term risk of cardiovascular mortality and disease based on a resting ECG alone. Using a large dataset of resting 12-lead ECGs collected at Stanford University Medical Center, we developed SEER, the Stanford Estimator of Electrocardiogram Risk. SEER predicts 5-year cardiovascular mortality with an area under the receiver operator characteristic curve (AUC) of 0.83 in a held-out test set at Stanford, and with AUCs of 0.78 and 0.83 respectively when independently evaluated at Cedars-Sinai Medical Center and Columbia University Irving Medical Center. SEER predicts 5-year atherosclerotic disease (ASCVD) with an AUC of 0.67, similar to the Pooled Cohort Equations for ASCVD Risk, while being only modestly correlated. When used in conjunction with the Pooled Cohort Equations, SEER accurately reclassified 16% of patients from low to moderate risk, uncovering a group with an actual average 9.9% 10-year ASCVD risk who would not have otherwise been indicated for statin therapy. SEER can also predict several other cardiovascular conditions such as heart failure and atrial fibrillation. Using only lead I of the ECG it predicts 5-year cardiovascular mortality with an AUC of 0.80. SEER, used alongside the Pooled Cohort Equations and other risk tools, can substantially improve cardiovascular risk stratification and aid in medical decision making.

The role of β-adrenergic stimulation in QT interval adaptation to heart rate during stress test

The adaptation lag of the QT interval after heart rate (HR) has been proposed as an arrhythmic risk marker. Most studies have quantified the QT adaptation lag in response to abrupt, step-like changes in HR induced by atrial pacing, in response to tilt test or during ambulatory recordings. Recent studies have introduced novel methods to quantify the QT adaptation lag to gradual, ramp-like HR changes in stress tests by evaluating the differences between the measured QT series and an estimated, memoryless QT series obtained from the instantaneous HR. These studies have observed the QT adaptation lag to progressively reduce when approaching the stress peak, with the underlying mechanisms being still unclear. This study analyzes the contribution of β-adrenergic stimulation to QT interval rate adaptation in response to gradual, ramp-like HR changes. We first quantify the QT adaptation lag in Coronary Artery Disease (CAD) patients undergoing stress test. To uncover the involved mechanisms, we use biophysically detailed computational models coupling descriptions of human ventricular electrophysiology and β-adrenergic signaling, from which we simulate ventricular action potentials and ECG signals. We characterize the adaptation of the simulated QT interval in response to the HR time series measured from each of the analyzed CAD patients. We show that, when the simulated ventricular tissue is subjected to a time-varying β-adrenergic stimulation pattern, with higher stimulation levels close to the stress peak, the simulated QT interval presents adaptation lags during exercise that are more similar to those measured from the patients than when subjected to constant β-adrenergic stimulation. During stress test recovery, constant and time-varying β-adrenergic stimulation patterns render similar adaptation lags, which are generally shorter than during exercise, in agreement with results from the patients. In conclusion, our findings support the role of time-varying β-adrenergic stimulation in contributing to QT interval adaptation to gradually increasing HR changes as those seen during the exercise phase of a stress test.

The 12-lead electrocardiogram as a biomarker of biological age

Background

We have demonstrated that a neural network is able to predict a person's age from the electrocardiogram (ECG) [artificial intelligence (AI) ECG age]. However, some discrepancies were observed between ECG-derived and chronological ages. We assessed whether the difference between AI ECG and chronological age (Age-Gap) represents biological ageing and predicts long-term outcomes.

Methods and results

We previously developed a convolutional neural network to predict chronological age from ECGs. In this study, we used the network to analyse standard digital 12-lead ECGs in a cohort of 25 144 subjects ≥30 years who had primary care outpatient visits from 1997 to 2003. Subjects with coronary artery disease, stroke, and atrial fibrillation were excluded. We tested whether Age-Gap was correlated with total and cardiovascular mortality. Of 25 144 subjects tested (54% females, 95% Caucasian) followed for 12.4 ± 5.3 years, the mean chronological age was 53.7 ± 11.6 years and ECG-derived age was 54.6 ± 11 years (R ² = 0.79, P < 0.0001). The mean Age-Gap was small at 0.88 ± 7.4 years. Compared to those whose ECG-derived age was within 1 standard deviation (SD) of their chronological age, patients with Age-Gap ≥1 SD had higher all-cause and cardiovascular disease (CVD) mortality. Conversely, subjects whose Age-Gap was ≤1 SD had lower all-cause and CVD mortality. Results were unchanged after adjusting for CVD risk factors and other survival influencing factors.

Conclusion

The difference between AI ECG and chronological age is an independent predictor of all-cause and cardiovascular mortality. Discrepancies between these possibly reflect disease independent biological ageing.

Prediction of mortality from 12-lead electrocardiogram voltage data using a deep neural network

The electrocardiogram (ECG) is a widely used medical test, consisting of voltage versus time traces collected from surface recordings over the heart¹. Here we hypothesized that a deep neural network (DNN) can predict an important future clinical event, 1-year all-cause mortality, from ECG voltage-time traces. By using ECGs collected over a 34-year period in a large regional health system, we trained a DNN with 1,169,662 12-lead resting ECGs obtained from 253,397 patients, in which 99,371 events occurred. The model achieved an area under the curve (AUC) of 0.88 on a held-out test set of 168,914 patients, in which 14,207 events occurred. Even within the large subset of patients (n = 45,285) with ECGs interpreted as 'normal' by a physician, the performance of the model in predicting 1-year mortality remained high (AUC = 0.85). A blinded survey of cardiologists demonstrated that many of the discriminating features of these normal ECGs were not apparent to expert reviewers. Finally, a Cox proportional-hazard model revealed a hazard ratio of 9.5 (P < 0.005) for the two predicted groups (dead versus alive 1 year after ECG) over a 25-year follow-up period. These results show that deep learning can add substantial prognostic information to the interpretation of 12-lead resting ECGs, even in cases that are interpreted as normal by physicians.

The effects of olive leaf extract and 28 days forced treadmill exercise on electrocardiographic parameters in rats

Background:

There is evidence that regular activity can prevent of cardiovascular diseases. There are many reports that exercise and the consumption of olive leaf extract (OLE) have a positive effect on cardiovascular parameters. This study was conducted to compare the effects of exercise and OLE alone and together on electrocardiographic parameters in rats.

Materials and Methods:

Male Sprague–Dawley rats were randomly divided into six groups (n = 8 rats in each): Control, exercise, OLE (100, 200, and 400 mg/kg, orally for 14 days), and exercise + OLE (200 mg/kg of extract, orally for 14 days). Exercise training in rats was performed using treadmill for 28 days (1 h/day). Electrophysiological parameters including heart rate, PR interval, QT interval, QT corrected (QTc), RR interval, QRS voltage, and duration were obtained from lead II electrocardiogram (ECG) recorded by a PowerLab system. Statistical evaluation was done by one-way analysis of variance followed by Fisher's least significant difference test and P < 0.05 was considered statistically significant.

Results:

The amounts of QT (P = 0.0009) and QTc interval (P = 0.0004), RR interval (P < 0.0001), QRS duration (P = 0.004), and QRS voltage (P = 0.003) in the exercise group were significantly higher than those of the control group. However, there were no significant differences in PR interval in comparison with the control group. Exercise (P < 0.0001) and OLE (400 mg/kg, P = 0.043) alone and both in combination (P = 0.007) reduced heart rate and increased the amount of QRS voltage (P = 0.003, P = 0.047, and P = 0.046, respectively) and RR interval (P < 0.0001, P = 0.046, and P = 0.0009, respectively).

Conclusion:

Results of this study indicated that administration of OLE alone and in combination with exercise has negative chronotropic and positive inotropic effects and also it can prevent of prolongation of QT and QTc interval induced by severe exercise.

Echocardiography and Electrocardiography Variables Correlate With the New York Heart Association classification: An Observational Study of Ischemic Cardiomyopathy Patients

The aim of our study was to determine whether combinations of ultrasound echocardiography (UCG) and electrocardiography (EKG) parameters correlated with the functional status of ischemic cardiomyopathy (ICM) patients according to the New York Heart Association (NYHA) classification system.We assessed 536 elderly Chinese ICM patients according to the NYHA criteria, which included 196 patients with type 2 diabetes mellitus (T2DM). All of the patients underwent UCG. Transmural dispersion of ventricular repolarization was examined using EKG. Cumulative odds logistic regression was performed to evaluate associations between NYHA class and the demographic, clinical, UCG, and EKG variables based on the odds ratio (OR) and 95% confidence interval (CI). A Pearson analysis was also performed to examine correlations between the NYHA classification and the UCG and EKG variables.Based on the NYHA assessment, 140, 147, 138, and 111 patients were identified as class I, II, III and IV, respectively. A comparison of UCG and EKG variables based on T2DM status showed that CO and Tp-e differed significantly between all NYHA classes (P < .05 for all), with values of each increasing with increasing NYHA class regardless of T2DM status. Multivariate logistic regression analysis showed that the disease course (OR: 1.30; 95% CI: 1.20-1.40), heart rate (OR: 1.16; 95% CI: 1.12-1.21), T wave peak to endpoint (Tp-e; OR: 1.22; 95% CI: 1.18-1.27), dispersion of the QT interval (OR: 0.98; 95% CI: 0.95-1.22), left ventricular fractional shortening (OR: 0.82; 95% CI: 0.78-0.87), cardiac output (CO; OR: 5.58; 95% CI: 3.08-10.13) were significantly associated with the NYHA class (P < .0001 for all). A Pearson correlation analysis showed that Tp-e (r = 0.75982, P < .0001), CO (r = 0.56072, P < .0001), and stroke volume (r = -0.14839, P = .0006) significantly correlated with the NYHA class.An index consisting of Tp-e and CO will be useful for corroborating the results of the NYHA assessment of ICM patients.

An Elective Course in Cardiovascular Electrophysiology for Pharmacy Learners

Objective. To implement an integrated, comprehensive, and learner-centered elective course focused at exposing learners to the interpretation of electrocardiograms and highlighting the mechanisms underlining the abnormal electrophysiological events. Design. Learners were presented with foundational information on the mechanisms underlying electrophysiological changes associated with the development of arrhythmias. They then discussed the interpretation of electrocardiogram (ECG) recordings and diagnosis of cardiovascular events. Teaching formats included "chalk-talk" and didactic sessions, case-based exercises providing hands-on evaluation of ECG recordings, and high-fidelity simulation presenting cases of arrhythmias. The course design emphasized critical thinking, learner engagement, and development of problem-solving skills. Learners were assessed by formal assignments, examinations, and in-class quizzes. Assessment. Learner comprehension of the material was assessed using cumulative examinations, in-class quizzes, assignments, and in-class presentations. Learner evaluations showed that the case-based discussions, practice ECGs, review tables, and illustrations enhanced course performance and retention of complex material. Conclusion. The elective course provided in-depth exposure to the mechanisms underlying electrophysiological aberrations resulting in arrhythmias. It gave learners an opportunity to learn the art of ECG interpretation and to apply their knowledge in simulated scenarios. As clinical teams adopt a multidisciplinary team approach to patient care, acquiring these skills enriches learner experiences and allows them to expand their role and professional opportunities as pharmacists.

Association of heart rate variability with arrhythmic events in patients with arrhythmogenic right ventricular cardiomyopathy/dysplasia

BACKGROUND

Arrhythmogenic right ventricular cardiomyopathy/dysplasia (ARVC/D) is associated with an increased risk of sudden cardiac death (SCD). Risk stratification of ARVC/D patients, however, remains an unresolved issue. In this study we investigated whether heart rate variability (HRV) can be helpful in identifying ARVC/D patients with increased risk of arrhythmic events.

METHODS AND RESULTS

We studied 30 consecutive patients (17 males; 45.4 ± 18 years) with ARVC/D, diagnosed according to guideline criteria; 15 patients (50%) had received an implantable cardioverter defibrillator (ICD) for primary SCD prevention. HRV was assessed on 24-h ECG Holter monitoring. The primary endpoint was the occurrence of major arrhythmic events (SCD, sustained ventricular tachycardia (VT), ICD therapy for sustained VT or ventricular fibrillation (VF)). During the follow-up period (19 ± 7 months), no deaths occurred, but 5 patients (17%) experienced arrhythmic events (4 VTs and 1 VF, all in the ICD group). All HRV parameters were significantly lower in patients with, compared with those without, arrhythmic events. Low-frequency amplitude was the most significant HRV variable associated with arrhythmic events in univariate Cox regression analysis (P=0.017), and was the only significant predictor of arrhythmic events in multivariable regression analysis (hazard ratio 0.88, P=0.047), together with unexplained syncope (hazard ratio 16.1, P=0.039).

CONCLUSIONS

Our data show that among ARVC/D patients HRV analysis might be helpful in identifying those with increased risk of major arrhythmic events.

High-frequency powers hidden within QRS complex as an additional predictor of lethal ventricular arrhythmias to ventricular late potential in post-myocardial infarction patients

BACKGROUND

Ventricular late potentials (VLPs) have been known to be a predictor of lethal ventricular arrhythmias (L-VAs); however, detection of other arrhythmogenic signals within the QRS complex remains obscure.

OBJECTIVE

The aim of this study was to evaluate whether abnormal intra-QRS high-frequency powers (IQHFP) within the QRS complex become a new predictor of L-VAs in addition to VLPs.

METHODS

Both 12-lead electrocardiograms (ECG) and VLPs were recorded from 142 subjects, including 37 patients without heart diseases, 97 patients post-myocardial infarction (MI), and 45 post-MI patients with L-VAs. Time-frequency analysis of ECG (leads V(1) or II) using wavelet transform with the Morlet function was performed. After the time-frequency powers were calculated, the ratios of the peak of signal power during the QRS complex in high-frequency bands against the peak power at 80 Hz (b/a ratio; P100, P150, P200, P250, or P300Hz/P80Hz) were measured. Abnormal IQHFP was defined when the b/a ratio exceeded the optimal cut-off values estimated by receiver-operator characteristic curves.

RESULTS

The combination of abnormal IQHFP appearing at 200, 250, and 300 Hz with positive VLPs increased the sensitivity for prediction of L-VAs from 53.3% by VLPs to 89.5%, and the negative predictive value from 74.7% by VLPs to 87.7%.

CONCLUSION

The combined use of VLPs and IQHFP hidden within the QRS complex improved the prediction of L-VAs in post-MI patients.

Genetic Polymorphisms as Risk Stratification Tool in Primary Preventive ICD Therapy

More and more implantable cardioverter-defibrillators (ICDs) are implanted as primary prevention of sudden cardiac death (SCD). However, major problem in practice is to identify high-risk patients for SCD. Different methods for noninvasive risk stratification do not have a sufficient positive or negative predictive value. Since current approaches lead to implantation of ICDs in a large number of patients who will never suffer an arrhythmic event and simultaneously patients still die of SCD who currently did not seem eligible for primary preventive ICD implantation, there is a need for additional tools for risk stratification. Epidemiological studies point to a hereditary risk of SCD. Different susceptibility of each person concerning arrhythmogenic events might be explained by genetic polymorphisms. By obtaining an individual “pattern” of polymorphisms of genes encoding for proteins which are important in arrhythmogenesis in one patient, risk stratification in primary prevention of SCD might by improved.

Risk Stratification for Sudden Cardiac Death after Acute Myocardial Infarction

Many patients who survive an acute myocardial infarction (AMI) remain at risk of recurrent cardiac events and sudden cardiac death after discharge, despite optimal medical treatment. Assessment of the degree of left ventricular dysfunction and residual myocardial ischaemia is useful to identify the patients at greatest risk. In addition, there is increasing evidence that a number of other cardiovascular tests can be used to detect autonomic dysfunction and myocardial substrate abnormalities postAMI that increase the risk of life-threatening ventricular arrhythmias. These investigations include ECG-based tests (signal averaged ECG and T-wave alternans), Holter-based recordings (heart rate variability and heart rate turbulence) and imaging techniques echocardiography and cardiac magnetic resonance), as well as invasive electrophysiological testing. This article reviews the current evidence for the use of these additional cardiac investigations among survivors of AMI to aid in their risk stratification for malignant ventricular arrhythmias and sudden cardiac death. Key words: Electrophysiological study, Holter recording, Non-invasive tests, Ventricular tachycardia

Electrocardiographic predictors of sudden cardiac death

Sudden cardiac death (SCD) is widespread and the most serious of the cardiac diseases, accounting for over half of cardiovascular mortality in adults in the United States, and nearly 1 in 3 of these patients does not report symptoms of cardiac disease before the sudden death. Quantifying the left ventricular ejection fraction is currently the best way to risk-stratify patients for SCD and identify those who are most likely to benefit from the insertion of an implantable cardiac defibrillator (ICD). The strategy of systemically placing ICDs in patients at risk of SCD is expensive and leads to substantial psychological hardship. However, noninvasive electrocardiographic indices of depolarization and repolarization may better identify patients who are at an increased risk of SCD. Therefore, developing an approach to identify electrocardiographic changes associated with the highest risk of arrhythmic death could markedly improve patient selection for ICD therapy. This report describes electrocardiographic parameters that may be useful in identifying patients at risk of SCD. The state of the science currently suggests that it is unlikely that a single electrocardiographic parameter will predict SCD, but rather a risk stratification algorithm based on a combination of electrocardiographic parameters may yield the best result.