P-Wave Area Predicts New Onset Atrial Fibrillation in Mitral Stenosis: A Machine Learning Approach

Atrial cardiomyopathy revisited-evolution of a concept: a clinical consensus statement of the European Heart Rhythm Association (EHRA) of the ESC, the Heart Rhythm Society (HRS), the Asian Pacific Heart Rhythm Society (APHRS), and the Latin American Heart Rhythm Society (LAHRS)

AIMS

The concept of "atrial cardiomyopathy" (AtCM) had been percolating through the literature since its first mention in 1972. Since then, publications using the term were sporadic until the decision was made to convene an expert working group with representation from four multinational arrhythmia organizations to prepare a consensus document on atrial cardiomyopathy in 2016 (EHRA/HRS/APHRS/SOLAECE expert consensus on atrial cardiomyopathies: definition, characterization, and clinical implication). Subsequently, publications on AtCM have increased progressively.

METHODS AND RESULTS

The present consensus document elaborates the 2016 AtCM document further to implement a simple AtCM staging system (AtCM stages 1-3) by integrating biomarkers, atrial geometry, and electrophysiological changes. However, the proposed AtCM staging needs clinical validation. Importantly, it is clearly stated that the presence of AtCM might serve as a substrate for the development of atrial fibrillation (AF) and AF may accelerates AtCM substantially, but AtCM per se needs to be viewed as a separate entity.

CONCLUSION

Thus, the present document serves as a clinical consensus statement of the European Heart Rhythm Association (EHRA) of the ESC, the Heart Rhythm Society (HRS), the Asian Pacific Heart Rhythm Society (APHRS), and the Latin American Heart Rhythm Society (LAHRS) to contribute to the evolution of the AtCM concept.

Role of the Electrocardiogram for Identifying the Development of Atrial Fibrillation

Atrial fibrillation (AF), a prevalent cardiac arrhythmia, is associated with increased morbidity and mortality worldwide. Stroke, the leading cause of serious disability in the United States, is among the important complications of this arrhythmia. Recent studies have demonstrated that certain clinical variables can be useful in the prediction of AF development in the future. The electrocardiogram (ECG) is a simple and cost-effective technology that is widely available in various healthcare settings. An emerging body of evidence has suggested that ECG tracings preceding the development of AF can be useful in predicting this arrhythmia in the future. Various variables on ECG especially different P wave parameters have been investigated in the prediction of new-onset AF and found to be useful. Several risk models were also introduced using these variables along with the patient's clinical data. However, current guidelines do not provide a clear consensus regarding implementing these prediction models in clinical practice for identifying patients at risk of AF. Also, the role of intensive screening via ECG or implantable devices based on this scoring system is unclear. The purpose of this review is to summarize AF and various related terminologies and explain the pathophysiology and electrocardiographic features of this tachyarrhythmia. We also discuss the predictive electrocardiographic features of AF, review some of the existing risk models and scoring system, and shed light on the role of monitoring device for screening purposes.

Electrocardiographic Predictors of Atrial Fibrillation

BACKGROUND

Atrial fibrillation (AF) is the most common pathological arrhythmia, and its complications lead to significant morbidity and mortality. However, patients with AF can often go undetected, especially if they are asymptomatic or have a low burden of paroxysms. Identification of those at high risk of AF development may help refine screening and management strategies.

METHODS

PubMed and Embase databases were systematically searched for studies looking at electrocardiographic predictors of AF from inception to August 2021.

RESULTS

A total of 115 studies were reported which examined a combination of atrial and ventricular parameters that could be electrocardiographic predictors of AF. Atrial predictors include conduction parameters, such as the PR interval, p-wave index and dispersion, and partial interatrial or advanced interatrial block, or morphological parameters, such as p-wave axis, amplitude and terminal force. Ventricular predictors include abnormalities in QRS amplitude, morphology or duration, QT interval duration, r-wave progression and ST segment, i.e., t-wave abnormalities.

CONCLUSIONS

There has been significant interest in electrocardiographic prediction of AF, especially in populations at high risk of atrial AF, such as those with an embolic stroke of undetermined source. This review highlights the breadth of possible predictive parameters, and possible pathological bases for the predictive role of each parameter are proposed.

P-wave durations from automated electrocardiogram analysis to predict atrial fibrillation and mortality in heart failure

BACKGROUND

P-wave indices have been used to predict incident atrial fibrillation (AF), stroke, and mortality. However, such indices derived from automated ECG measurements have not been explored for their predictive values in heart failure (HF). We investigated whether automated P-wave indices can predict adverse outcomes in HF.

METHODS

This study included consecutive Chinese patients admitted to a single tertiary centre, presenting with HF but without prior AF, and with at least one baseline ECG, between 1 January 2010 and 31 December 2016, with last follow-up of 31 December 2019.

RESULTS

A total of 2718 patients were included [median age: 77.4, interquartile range (IQR): (66.9-84.3) years; 47.9 males]. After a median follow-up of 4.8 years (IQR: 1.9-9.0 years), 1150 patients developed AF (8.8/year), 339 developed stroke (2.6/year), 563 developed cardiovascular mortality (4.3/year), and 1972 had all-cause mortality (15.1/year). Compared with 101-120 ms as a reference, maximum P-wave durations predicted new-onset AF at ≤90 ms [HR: 1.17(1.11, 1.50), P < 0.01], 131-140 ms [HR: 1.29(1.09, 1.54), P < 0.001], and ≥141 ms [HR: 1.52(1.32, 1.75), P < 0.001]. Similarly, they predicted cardiovascular mortality at ≤90 ms [HR: 1.50(1.08, 2.06), P < 0.001] or ≥141 ms [HR: 1.18(1.15, 1.45), P < 0.001], and all-cause mortality at ≤90 ms [HR: 1.26(1.04, 1.51), P < 0.001], 131-140 ms [HR: 1.15(1.01, 1.32), P < 0.01], and ≥141 ms [HR: 1.31(1.18, 1.46), P < 0.001]. These remained significant after adjusting for significant demographics, past co-morbidities, P-wave dispersion, and maximum P-wave amplitude.

CONCLUSIONS

Extreme values of maximum P-wave durations (≤90 ms and ≥141 ms) were significant predictors of new-onset AF, cardiovascular mortality, and all-cause mortality.

Predicting Stroke and Mortality in Mitral Regurgitation: A Machine Learning Approach

We hypothesized that an interpretable gradient boosting machine (GBM) model considering comorbidities, P-wave and echocardiographic measurements, can better predict mortality and cerebrovascular events in mitral regurgitation (MR). Patients from a tertiary center were analyzed. The GBM model was used as an interpretable statistical approach to identify the leading indicators of high-risk patients with either outcome of CVAs and all-cause mortality. A total of 706 patients were included. GBM analysis showed that age, systolic blood pressure, diastolic blood pressure, plasma albumin levels, mean P-wave duration (PWD), MR regurgitant volume, left ventricular ejection fraction (LVEF), left atrial dimension at end-systole (LADs), velocity-time integral (VTI) and effective regurgitant orifice were significant predictors of TIA/stroke. Age, sodium, urea and albumin levels, platelet count, mean PWD, LVEF, LADs, left ventricular dimension at end systole (LVDs) and VTI were significant predictors of all-cause mortality. The GBM demonstrates the best predictive performance in terms of precision, sensitivity c-statistic and F1-score compared to logistic regression, decision tree, random forest, support vector machine, and artificial neural networks. Gradient boosting model incorporating clinical data from different investigative modalities significantly improves risk prediction performance and identify key indicators for outcome prediction in MR.

Long-term single-lead electrocardiogram monitoring to detect new-onset postoperative atrial fibrillation in patients after cardiac surgery

Background

Postoperative atrial fibrillation (POAF) is often associated with serious complications. In this study, we collected long-term single-lead electrocardiograms (ECGs) of patients with preoperative sinus rhythm to build statistical models and machine learning models to predict POAF.

Methods

All patients with preoperative sinus rhythm who underwent cardiac surgery were enrolled and we collected long-term ECG data 24 h before surgery and 7 days after surgery by single-lead ECG. The patients were divided into a POAF group a no-POAF group. A clinical model and a clinical + ECG model were constructed. The ECG parameters were designed and support vector machine (SVM) was selected to build a machine learning model and evaluate its prediction efficiency.

Results

A total of 100 patients were included. The detection rate of POAF in long-term ECG monitoring was 31% and that in conventional monitoring was 19%. We calculated 7 P-wave parameters, Pmax (167 ± 31 ms vs. 184 ± 37 ms, P = 0.018), Pstd (15 ± 7 vs. 19 ± 11, P = 0.031), and PWd (62 ± 28 ms vs. 80 ± 35 ms, P = 0.008) were significantly different. The AUC of the clinical model (sex, age, LA diameter, GFR, mechanical ventilation time) was 0.86. Clinical + ECG model (sex, age, LA diameter, GFR, mechanical ventilation time, Pmax, Pstd, PWd), AUC was 0.89. In the machine learning model, the accuracy (Ac) of the train set and test set was above 80 and 60%, respectively.

Conclusion

Long-term ECG monitoring could significantly improve the detection rate of POAF. The clinical + ECG model and the machine learning model based on P-wave parameters can predict POAF.

P Wave Parameters and Indices: A Critical Appraisal of Clinical Utility, Challenges, and Future Research-A Consensus Document Endorsed by the International Society of Electrocardiology and the International Society for Holter and Noninvasive Electrocardiology

Atrial cardiomyopathy, characterized by abnormalities in atrial structure and function, is associated with increased risk of adverse cardiovascular and neurocognitive outcomes, independent of atrial fibrillation. There exists a critical unmet need for a clinical tool that is cost-effective, easy to use, and that can diagnose atrial cardiomyopathy. P wave parameters (PWPs) reflect underlying atrial structure, size, and electrical activation; alterations in these factors manifest as abnormalities in PWPs that can be readily ascertained from a standard 12-lead ECG and potentially be used to aid clinical decision-making. PWPs include P wave duration, interatrial block, P wave terminal force in V₁, P wave axis, P wave voltage, P wave area, and P wave dispersion. PWPs can be combined to yield an index (P wave index), such as the morphology-voltage-P-wave duration ECG risk score. Abnormal PWPs have been shown in population-based cohort studies to be independently associated with higher risks of atrial fibrillation, ischemic stroke, sudden cardiac death, and dementia. Additionally, PWPs, either individually or in combination (as a P wave index), have been reported to enhance prediction of atrial fibrillation or ischemic stroke. To facilitate translation of PWPs to routine clinical practice, additional work is needed to standardize measurement of PWPs (eg, via semiautomated or automated measurement), confirm their reliability and predictive value, leverage novel approaches (eg, wavelet analysis of P waves and machine learning algorithms), and finally, define the risk-benefit ratio of specific interventions in high-risk individuals. Our ultimate goal is to repurpose the ubiquitous 12-lead ECG to advance the study, diagnosis, and treatment of atrial cardiomyopathy, thus overcoming critical challenges in prevention of cardiovascular disease and dementia.

ECG Predictors for New-Onset Atrial Fibrillation Within a Year After Radiofrequency Ablation of Counterclockwise-Rotating Atrial Flutter

Background: New-onset atrial fibrillation (AF) after ablation of typical atrial flutter (AFL) is not rare. This study aimed to investigate the predictive value of electrocardiographic parameters on new-onset AF post-typical AFL ablation.

Methods: A total of 158 consecutive patients (79.1% males, mean age 57.8 ± 14.3 years) with typical AFL were enrolled between January 2012 and August 2017 in this single-center study. Patients with a history of AF before ablation were excluded. ECGs during sinus rhythm (SR) and AFL were collected. The duration of the negative component of flutter wave in lead II (D_FNII), proportion of the D_FNII of the total circle length of AFL (D_FNII%), amplitude of the negative component of flutter wave in lead II (A_FNII), duration (D_PNV1), and amplitude (A_PNV1) of negative component of the P wave in lead V1, and P wave duration in lead II (D_PII) during sinus rhythm were measured.

Results: During a median follow-up of 26.9 ± 11.8 months, 22 cases (13.9%) developed new-onset AF. D_FNII was significantly longer in patients with new-onset AF compared to patients without AF (114.7 ± 29.6 ms vs. 82.7 ± 12.8 ms, p < 0.0001). A_FNII was significantly lower (0.118 ± 0.034 mV vs. 0.168 ± 0.051 mV, p < 0.0001), D_PII (144.21 ± 23.77 ms vs. 111.46 ± 14.19 ms, p < 0.0001), and D_PNV1 was significantly longer (81.07 ± 16.87 ms vs. 59.86 ± 14.42 ms, p < 0.0001) in patients with new-onset AF. In the multivariate analysis, D_FNII [odds ratio (OR), 1.428; 95% CI, 1.039–1.962; p = 0.028] and D_PII (OR, 1.429; 95% CI, 1.046–1.953; p = 0.025) were found to be independently associated with new-onset AF after typical AFL ablation.

Conclusion: Parameters representing left atrial activation time under both the SR and AFL were independently associated with new-onset AF post-typical AFL ablation and may be useful in risk prediction, which needs to be confirmed by further prospective studies.

How machine learning is impacting research in atrial fibrillation: implications for risk prediction and future management

There has been an exponential growth of artificial intelligence (AI) and machine learning (ML) publications aimed at advancing our understanding of atrial fibrillation (AF), which has been mainly driven by the confluence of two factors: the advances in deep neural networks (DeepNNs) and the availability of large, open access databases. It is observed that most of the attention has centred on applying ML for dvsetecting AF, particularly using electrocardiograms (ECGs) as the main data modality. Nearly a third of them used DeepNNs to minimize or eliminate the need for transforming the ECGs to extract features prior to ML modelling; however, we did not observe a significant advantage in following this approach. We also found a fraction of studies using other data modalities, and others centred in aims, such as risk prediction, AF management, and others. From the clinical perspective, AI/ML can help expand the utility of AF detection and risk prediction, especially for patients with additional comorbidities. The use of AI/ML for detection and risk prediction into applications and smart mobile health (mHealth) technology would enable ‘real time’ dynamic assessments. AI/ML could also adapt to treatment changes over time, as well as incident risk factors. Incorporation of a dynamic AI/ML model into mHealth technology would facilitate ‘real time’ assessment of stroke risk, facilitating mitigation of modifiable risk factors (e.g. blood pressure control). Overall, this would lead to an improvement in clinical care for patients with AF.

Non-Invasive and Quantitative Estimation of Left Atrial Fibrosis Based on P Waves of the 12-Lead ECG-A Large-Scale Computational Study Covering Anatomical Variability

The arrhythmogenesis of atrial fibrillation is associated with the presence of fibrotic atrial tissue. Not only fibrosis but also physiological anatomical variability of the atria and the thorax reflect in altered morphology of the P wave in the 12-lead electrocardiogram (ECG). Distinguishing between the effects on the P wave induced by local atrial substrate changes and those caused by healthy anatomical variations is important to gauge the potential of the 12-lead ECG as a non-invasive and cost-effective tool for the early detection of fibrotic atrial cardiomyopathy to stratify atrial fibrillation propensity. In this work, we realized 54,000 combinations of different atria and thorax geometries from statistical shape models capturing anatomical variability in the general population. For each atrial model, 10 different volume fractions (0-45%) were defined as fibrotic. Electrophysiological simulations in sinus rhythm were conducted for each model combination and the respective 12-lead ECGs were computed. P wave features (duration, amplitude, dispersion, terminal force in V1) were extracted and compared between the healthy and the diseased model cohorts. All investigated feature values systematically in- or decreased with the left atrial volume fraction covered by fibrotic tissue, however value ranges overlapped between the healthy and the diseased cohort. Using all extracted P wave features as input values, the amount of the fibrotic left atrial volume fraction was estimated by a neural network with an absolute root mean square error of 8.78%. Our simulation results suggest that although all investigated P wave features highly vary for different anatomical properties, the combination of these features can contribute to non-invasively estimate the volume fraction of atrial fibrosis using ECG-based machine learning approaches.

Automated Electrocardiogram Analysis Identifies Novel Predictors of Ventricular Arrhythmias in Brugada Syndrome

Background: Patients suffering from Brugada syndrome (BrS) are at an increased risk of life-threatening ventricular arrhythmias. Whilst electrocardiographic (ECG) variables have been used for risk stratification with varying degrees of success, automated measurements have not been tested for their ability to predict adverse outcomes in BrS. Methods: BrS patients presenting in a single tertiary center between 2000 and 2018 were analyzed retrospectively. ECG variables on vector magnitude, axis, amplitude and duration from all 12 leads were determined. The primary endpoint was spontaneous ventricular tachycardia/ventricular fibrillation (VT/VF) on follow-up. Results: This study included 83 patients [93% male, median presenting age: 56 (41-66) years old, 45% type 1 pattern] with 12 developing the primary endpoint (median follow-up: 75 (Q1-Q3: 26-114 months). Cox regression showed that QRS frontal axis > 70.0 degrees, QRS horizontal axis > 57.5 degrees, R-wave amplitude (lead I) <0.67 mV, R-wave duration (lead III) > 50.0 ms, S-wave amplitude (lead I) < -0.144 mV, S-wave duration (lead aVL) > 35.5 ms, QRS duration (lead V3) > 96.5 ms, QRS area in lead I < 0.75 Ashman units, ST slope (lead I) > 31.5 deg, T-wave area (lead V1) < -3.05 Ashman units and PR interval (lead V2) > 157 ms were significant predictors. A weighted score based on dichotomized values provided good predictive performance (hazard ratio: 1.59, 95% confidence interval: 1.27-2.00, P-value<0.0001, area under the curve: 0.84). Conclusions: Automated ECG analysis revealed novel risk markers in BrS. These markers should be validated in larger prospective studies.

Multi-modality machine learning approach for risk stratification in heart failure with left ventricular ejection fraction ≤ 45

AIMS

Heart failure (HF) involves complex remodelling leading to electrical and mechanical dysfunction. We hypothesized that machine learning approaches incorporating data obtained from different investigative modalities including atrial and ventricular measurements from electrocardiography and echocardiography, blood inflammatory marker [neutrophil-to-lymphocyte ratio (NLR)], and prognostic nutritional index (PNI) will improve risk stratification for adverse outcomes in HF compared to logistic regression.

METHODS AND RESULTS

Consecutive Chinese patients referred to our centre for transthoracic echocardiography and subsequently diagnosed with HF, between 1 January 2010 and 31 December 2016, were included in this study. Two machine learning techniques, multilayer perceptron and multi-task learning, were compared with logistic regression for their ability to predict incident atrial fibrillation (AF), transient ischaemic attack (TIA)/stroke, and all-cause mortality. This study included 312 HF patients [mean age: 64 (55-73) years, 75% male]. There were 76 cases of new-onset AF, 62 cases of incident TIA/stroke, and 117 deaths during follow-up. Univariate analysis revealed that age, left atrial reservoir strain (LARS) and contractile strain (LACS) were significant predictors of new-onset AF. Age and smoking predicted incident stroke. Age, hypertension, type 2 diabetes mellitus, chronic kidney disease, mitral or aortic regurgitation, P-wave terminal force in V1, the presence of partial inter-atrial block, left atrial diameter, ejection fraction, global longitudinal strain, serum creatinine and albumin, high NLR, low PNI, and LARS and LACS predicted all-cause mortality. Machine learning techniques achieved better prediction performance than logistic regression.

CONCLUSIONS

Multi-modality assessment is important for risk stratification in HF. A machine learning approach provides additional value for improving outcome prediction.

Multi-parametric system for risk stratification in mitral regurgitation: A multi-task Gaussian prediction approach

BACKGROUND

We hypothesized that a multi-parametric approach incorporating medical comorbidity information, electrocardiographic P-wave indices, echocardiographic assessment, neutrophil-to-lymphocyte ratio (NLR) and prognostic nutritional index (PNI) calculated from laboratory data can improve risk stratification in mitral regurgitation (MR).

METHODS

Patients diagnosed with mitral regurgitation between 1 March 2005 and 30 October 2018 from a single centre were retrospectively analysed. Outcomes analysed were incident atrial fibrillation (AF), transient ischemic attack (TIA)/stroke and mortality.

RESULTS

This study cohort included 706 patients, of whom 171 had normal inter-atrial conduction, 257 had inter-atrial block (IAB) and 266 had AF at baseline. Logistic regression analysis showed that age, hypertension and mean P-wave duration (PWD) were significant predictors of new-onset AF. Low left ventricular ejection fraction (LVEF), abnormal P-wave terminal force in V1 (PTFV1) predicted TIA/stroke. Age, smoking, hypertension, diabetes mellitus, hypercholesterolaemia, ischemic heart disease, secondary mitral regurgitation, urea, creatinine, NLR, PNI, left atrial diameter (LAD), left ventricular end-diastolic dimension, LVEF, pulmonary arterial systolic pressure, IAB, baseline AF and heart failure predicted all-cause mortality. A multi-task Gaussian process learning model demonstrated significant improvement in risk stratification compared to logistic regression and a decision tree method.

CONCLUSIONS

A multi-parametric approach incorporating multi-modality clinical data improves risk stratification in mitral regurgitation. Multi-task machine learning can significantly improve overall risk stratification performance.