Sex-related electrocardiographic differences in patients with different types of atrial fibrillation: Results from the SWISS-AF study

Electrocardiographic parameters and the risk of new-onset atrial fibrillation in the general population: the Rotterdam Study

AIMS

We aimed to assess the (shape of the) association and sex differences in the link between electrocardiographic parameters and new-onset atrial fibrillation (AF).

METHODS AND RESULTS

A total of 12 212 participants free of AF at baseline from the population-based Rotterdam Study were included. Up to five repeated measurements of electrocardiographic parameters including PR, QRS, QT, QT corrected for heart rate (QTc), JT, RR interval, and heart rate were assessed at baseline and follow-up examinations. Cox proportional hazards- and joint models, adjusted for cardiovascular risk factors, were used to determine the (shape of the) association between baseline and longitudinal electrocardiographic parameters with new-onset AF. Additionally, we evaluated potential sex differences. During a median follow-up of 9.3 years, 1282 incident AF cases occurred among 12 212 participants (mean age 64.9 years, 58.2% women). Penalized cubic splines revealed that associations between baseline electrocardiographic measures and risk of new-onset AF were generally U- and N-shaped. Sex differences in terms of the shape of the various associations were most apparent for baseline PR, QT, QTc, RR interval, and heart rate in relation to new-onset AF. Longitudinal measures of higher PR interval [fully adjusted hazard ratio (HR), 95% confidence interval (CI), 1.43, 1.02-2.04, P = 0.0393] and higher QTc interval (fully adjusted HR, 95% CI, 5.23, 2.18-12.45, P = 0.0002) were significantly associated with new-onset AF, in particular in men.

CONCLUSION

Associations of baseline electrocardiographic measures and risk of new-onset AF were mostly U- and N-shaped. Longitudinal electrocardiographic measures of PR and QTc interval were significantly associated with new-onset AF, in particular among men.

Perspectives on Sex- and Gender-Specific Prediction of New-Onset Atrial Fibrillation by Leveraging Big Data

Atrial fibrillation (AF), the most common sustained cardiac arrhythmia, has a large impact on quality of life and is associated with increased risk of hospitalization, morbidity, and mortality. Over the past two decades advances regarding the clinical epidemiology and management of AF have been established. Moreover, sex differences in the prevalence, incidence, prediction, pathophysiology, and prognosis of AF have been identified. Nevertheless, AF remains to be a complex and heterogeneous disorder and a comprehensive sex- and gender-specific approach to predict new-onset AF is lacking. The exponential growth in various sources of big data such as electrocardiograms, electronic health records, and wearable devices, carries the potential to improve AF risk prediction. Leveraging these big data sources by artificial intelligence (AI)-enabled approaches, in particular in a sex- and gender-specific manner, could lead to substantial advancements in AF prediction and ultimately prevention. We highlight the current status, premise, and potential of big data to improve sex- and gender-specific prediction of new-onset AF.

Extended ECG Improves Classification of Paroxysmal and Persistent Atrial Fibrillation Based on P- and f-Waves

Background

The standard 12-lead ECG has been shown to be of value in characterizing atrial conduction properties. The added value of extended ECG recordings (longer recordings from more sites) has not been systematically explored yet.

Objective

The aim of this study is to employ an extended ECG to identify characteristics of atrial electrical activity related to paroxysmal vs. persistent atrial fibrillation (AF).

Methods

In 247 participants scheduled for AF ablation, an extended ECG was recorded (12 standard plus 3 additional leads, 5 min recording, no filtering). For patients presenting in sinus rhythm (SR), the signal-averaged P-wave and the spatiotemporal P-wave variability was analyzed. For patients presenting in AF, f-wave properties in the QRST (the amplitude complex of the ventricular electrical activity: Q-, R-, S-, and T-wave)-canceled ECG were determined.

Results

Significant differences between paroxysmal (N = 152) and persistent patients with AF (N = 95) were found in several P-wave and f-wave parameters, including parameters that can only be calculated from an extended ECG. Furthermore, a moderate, but significant correlation was found between echocardiographic parameters and P-wave and f-wave parameters. There was a moderate correlation of left atrial (LA) diameter with P-wave energy duration (r = 0.317, p < 0.001) and f-wave amplitude in lead A3 (r = -0.389, p = 0.002). The AF-type classification performance significantly improved when parameters calculated from the extended ECG were taken into account [area under the curve (AUC) = 0.58, interquartile range (IQR) 0.50-0.64 for standard ECG parameters only vs. AUC = 0.76, IQR 0.70-0.80 for extended ECG parameters, p < 0.001].

Conclusion

The P- and f-wave analysis of extended ECG configurations identified specific ECG features allowing improved classification of paroxysmal vs. persistent AF. The extended ECG significantly improved AF-type classification in our analyzed data as compared to a standard 10-s 12-lead ECG. Whether this can result in a better clinical AF type classification warrants further prospective study.

FASSNet: fast apnea syndrome screening neural network based on single-lead electrocardiogram for wearable devices

Objective. Sleep apnea (SA) is a chronic condition that fragments sleep and results in intermittent hypoxemia, which in long run leads to cardiovascular diseases like stroke. Diagnosis of SA through polysomnography is costly, inconvenient, and has long waiting list. Wearable devices provide a low-cost solution to the ambulatory detection of SA syndrome for undiagnosed patients. One of the wearables are the ones based on minute-by-minute analysis of single-lead electrocardiogram (ECG) signal. Processing ECG segments online at wearables contributes to memory conservation and privacy protection in long-term SA monitoring, and light-weight models are required due to stringent computation resource.Approach.We propose fast apnea syndrome screening neural network (FASSNet), an effective end-to-end neural network to perform minute-apnea event detection. Low-frequency components of filtered ECG spectrogram are selected as input. The model initially processes the spectrogram via convolution blocks. Bidirectional long-short-term memory blocks are used along the frequency axis to complement position information of frequency bands. Layer normalisation is implemented to retain in-epoch information since apnea periods have variable lengths. Experiments were carried out on 70 recordings of Apnea-ECG database, where each 60 s ECG segment is manually labelled as an apnea or normal minute by technician. Both ten-fold and patient-agnostic validation protocols are adopted.Main results.FASSNet is light-weighted, since its value of model parameters and multiply accumulates are 0.06% and 28.33% of those of an AlexNet benchmark, respectively. Meanwhile, FASSNet achieves an accuracy of 87.09%, a sensitivity of 77.96%, a specificity of 91.74%, and an F1 score of 81.61% in apnea event detection. Its accuracy of diagnosing SA syndrome severity exceeds 90% under the patient-agnostic protocol.Significance:FASSNet is a computationally efficient and accurate neural network for wearables to detect SA events and estimate SA severity based on minute-level diagnosis.

Atrial fibrillation risk prediction from the 12-lead electrocardiogram using digital biomarkers and deep representation learning

Aims

This study aims to assess whether information derived from the raw 12-lead electrocardiogram (ECG) combined with clinical information is predictive of atrial fibrillation (AF) development.

Methods and results

We use a subset of the Telehealth Network of Minas Gerais (TNMG) database consisting of patients that had repeated 12-lead ECG measurements between 2010 and 2017 that is 1 130 404 recordings from 415 389 unique patients. Median and interquartile of age for the recordings were 58 (46-69) and 38% of the patients were males. Recordings were assigned to train-validation and test sets in an 80:20% split which was stratified by class, age and gender. A random forest classifier was trained to predict, for a given recording, the risk of AF development within 5 years. We use features obtained from different modalities, namely demographics, clinical information, engineered features, and features from deep representation learning. The best model performance on the test set was obtained for the model combining features from all modalities with an area under the receiver operating characteristic curve (AUROC) = 0.909 against the best single modality model which had an AUROC = 0.839.

Conclusion

Our study has important clinical implications for AF management. It is the first study integrating feature engineering, deep learning, and Electronic medical record system (EMR) metadata to create a risk prediction tool for the management of patients at risk of AF. The best model that includes features from all modalities demonstrates that human knowledge in electrophysiology combined with deep learning outperforms any single modality approach. The high performance obtained suggest that structural changes in the 12-lead ECG are associated with existing or impending AF.

Short P-Wave Duration is a Marker of Higher Rate of Atrial Fibrillation Recurrences after Pulmonary Vein Isolation: New Insights into the Pathophysiological Mechanisms Through Computer Simulations

Background Short ECG P-wave duration has recently been demonstrated to be associated with higher risk of atrial fibrillation (AF). The aim of this study was to assess the rate of AF recurrence after pulmonary vein isolation in patients with a short P wave, and to mechanistically elucidate the observation by computer modeling. Methods and Results A total of 282 consecutive patients undergoing a first single-pulmonary vein isolation procedure for paroxysmal or persistent AF were included. Computational models studied the effect of adenosine and sodium conductance on action potential duration and P-wave duration (PWD). About 16% of the patients had a PWD of 110 ms or shorter (median PWD 126 ms, interquartile range, 115 ms-138 ms; range, 71 ms-180 ms). At Cox regression, PWD was significantly associated with AF recurrence (P=0.012). Patients with a PWD <110 ms (hazard ratio [HR], 2.20; 95% CI, 1.24-3.88; P=0.007) and patients with a PWD ≥140 (HR, 1.87, 95% CI, 1.06-3.30; P=0.031) had a nearly 2-fold increase in risk with respect to the other group. In the computational model, adenosine yielded a significant reduction of action potential duration 90 (52%) and PWD (7%). An increased sodium conductance (up to 200%) was robustly accompanied by an increase in conduction velocity (26%), a reduction in action potential duration 90 (28%), and PWD (22%). Conclusions One out of 5 patients referred for pulmonary vein isolation has a short PWD which was associated with a higher rate of AF after the index procedure. Computer simulations suggest that shortening of atrial action potential duration leading to a faster atrial conduction may be the cause of this clinical observation.