Deep Learning Applied to Electrocardiogram Interpretation

Optimization of Arrhythmia-based ECG-lead Selection for Computer-interpreted Heart Rhythm Classification

The 12-lead ECG only has 8 independent ECG leads, which leads to diagnostic redundancy when using all 12 leads for heart arrhythmias classification. We have previously developed a deep learning (DL)-based computer-interpreted ECG (CIE) approach to identify an optimal 4-lead ECG subset for classifying heart arrhythmias. However, the clinical diagnostic criteria of cardiac arrhythmia types are often lead-specific, so this study is going to explore the selection of arrhythmia-based ECG-lead subsets rather than one general optimal ECG-lead subset, which could improve the classification performance for the CIE. The DL-based CIE model previously developed was used to learn 4 common types of heart arrhythmias (LBBB, RBBB, AF, and I-AVB) for identifying corresponding optimal ECG-lead subsets. A public dataset that splits into training (approx. 70%), validation (approx. 15%), and test (approx. 15%) sets from the PhysioNet Cardiology Challenge 2020 was used to explore the study. The results demonstrated that the DL-based CIE model identified an optimal ECG-lead subset for each arrhythmia: I, II, aVR, aVL, V1, V3, and V5 for I-AVB; I, II, aVR, and V3 for AF; I, II, aVR, aVF, V1, V3, and V4 for LBBB; and I, II, III, aVR, V1, V4, and V6 for RBBB. For each arrhythmia classification, the DL-based CIE model using the optimal ECG-lead subset significantly outperformed the model using the full 12-lead ECG set on the validation set and on the external test dataset.The results support the hypothesis that using an optimal ECG-lead subset instead of the full 12-lead ECG set can improve the classification performance of a specific arrhythmia when using the DL-based CIE approach.Clinical Relevance- Using an arrhythmia-based optimal ECG-lead subset, the classification performance of a deep-learning-based model can be achieved without loss of accuracy in comparison with the full 12-lead set (p<0.05).

Improving localization accuracy for non-invasive automated early left ventricular origin localization approach

Background: We previously developed a non-invasive approach to localize the site of early left ventricular activation origin in real time using 12-lead ECG, and to project the predicted site onto a generic LV endocardial surface using the smallest angle between two vectors algorithm (SA). Objectives: To improve the localization accuracy of the non-invasive approach by utilizing the K-nearest neighbors algorithm (KNN) to reduce projection errors. Methods: Two datasets were used. Dataset #1 had 1012 LV endocardial pacing sites with known coordinates on the generic LV surface and corresponding ECGs, while dataset #2 included 25 clinically-identified VT exit sites and corresponding ECGs. The non-invasive approach used "population" regression coefficients to predict the target coordinates of a pacing site or VT exit site from the initial 120-m QRS integrals of the pacing site/VT ECG. The predicted site coordinates were then projected onto the generic LV surface using either the KNN or SA projection algorithm. Results: The non-invasive approach using the KNN had a significantly lower mean localization error than the SA in both dataset #1 (9.4 vs. 12.5 mm, p < 0.05) and dataset #2 (7.2 vs. 9.5 mm, p < 0.05). The bootstrap method with 1,000 trials confirmed that using KNN had significantly higher predictive accuracy than using the SA in the bootstrap assessment with the left-out sample (p < 0.05). Conclusion: The KNN significantly reduces the projection error and improves the localization accuracy of the non-invasive approach, which shows promise as a tool to identify the site of origin of ventricular arrhythmia in non-invasive clinical modalities.

The Atrium in Atrial Fibrillation - A Clinical Review on How to Manage Atrial Fibrotic Substrates

Atrial fibrillation (AF) is the most common sustained arrhythmia in the population and is associated with a significant clinical and economic burden. Rigorous assessment of the presence and degree of an atrial arrhythmic substrate is essential for determining treatment options, predicting long-term success after catheter ablation, and as a substrate critical in the pathophysiology of atrial thrombogenesis. Catheter ablation of AF has developed into an essential rhythm-control strategy. Nowadays is one of the most common cardiac ablation procedures performed worldwide, with its success inversely related to the extent of atrial structural disease. Although atrial substrate evaluation remains complex, several diagnostic resources allow for a more comprehensive assessment and quantification of the extent of left atrial structural remodeling and the presence of atrial fibrosis. In this review, we summarize the current knowledge on the pathophysiology, etiology, and electrophysiological aspects of atrial substrates promoting the development of AF. We also describe the risk factors for its development and how to diagnose its presence using imaging, electrocardiograms, and electroanatomic voltage mapping. Finally, we discuss recent data regarding fibrosis biomarkers that could help diagnose atrial fibrotic substrates.

Optimal ECG-lead selection increases generalizability of deep learning on ECG abnormality classification

Deep learning (DL) has achieved promising performance in detecting common abnormalities from the 12-lead electrocardiogram (ECG). However, diagnostic redundancy exists in the 12-lead ECG, which could impose a systematic overfitting on DL, causing poor generalization. We, therefore, hypothesized that finding an optimal lead subset of the 12-lead ECG to eliminate the redundancy would help improve the generalizability of DL-based models. In this study, we developed and evaluated a DL-based model that has a feature extraction stage, an ECG-lead subset selection stage and a decision-making stage to automatically interpret multiple common ECG abnormality types. The data analysed in this study consisted of 6877 12-lead ECG recordings from CPSC 2018 (labelled as normal rhythm or eight types of ECG abnormalities, split into training (approx. 80%), validation (approx. 10%) and test (approx. 10%) sets) and 3998 12-lead ECG recordings from PhysioNet/CinC 2020 (labelled as normal rhythm or four types of ECG abnormalities, used as external text set). The ECG-lead subset selection module was introduced within the proposed model to efficiently constrain model complexity. It detected an optimal 4-lead ECG subset consisting of leads II, aVR, V1 and V4. The proposed model using the optimal 4-lead subset significantly outperformed the model using the complete 12-lead ECG on the validation set and on the external test dataset. The results demonstrated that our proposed model successfully identified an optimal subset of 12-lead ECG; the resulting 4-lead ECG subset improves the generalizability of the DL model in ECG abnormality interpretation. This study provides an outlook on what channels are necessary to keep and which ones may be ignored when considering an automated detection system for cardiac ECG abnormalities. This article is part of the theme issue 'Advanced computation in cardiovascular physiology: new challenges and opportunities'.

Combined In-silico and Machine Learning Approaches Toward Predicting Arrhythmic Risk in Post-infarction Patients

Background: Remodeling due to myocardial infarction (MI) significantly increases patient arrhythmic risk. Simulations using patient-specific models have shown promise in predicting personalized risk for arrhythmia. However, these are computationally- and time- intensive, hindering translation to clinical practice. Classical machine learning (ML) algorithms (such as K-nearest neighbors, Gaussian support vector machines, and decision trees) as well as neural network techniques, shown to increase prediction accuracy, can be used to predict occurrence of arrhythmia as predicted by simulations based solely on infarct and ventricular geometry. We present an initial combined image-based patient-specific in silico and machine learning methodology to assess risk for dangerous arrhythmia in post-infarct patients. Furthermore, we aim to demonstrate that simulation-supported data augmentation improves prediction models, combining patient data, computational simulation, and advanced statistical modeling, improving overall accuracy for arrhythmia risk assessment.

Methods: MRI-based computational models were constructed from 30 patients 5 days post-MI (the “baseline” population). In order to assess the utility biophysical model-supported data augmentation for improving arrhythmia prediction, we augmented the virtual baseline patient population. Each patient ventricular and ischemic geometry in the baseline population was used to create a subfamily of geometric models, resulting in an expanded set of patient models (the “augmented” population). Arrhythmia induction was attempted via programmed stimulation at 17 sites for each virtual patient corresponding to AHA LV segments and simulation outcome, “arrhythmia,” or “no-arrhythmia,” were used as ground truth for subsequent statistical prediction (machine learning, ML) models. For each patient geometric model, we measured and used choice data features: the myocardial volume and ischemic volume, as well as the segment-specific myocardial volume and ischemia percentage, as input to ML algorithms. For classical ML techniques (ML), we trained k-nearest neighbors, support vector machine, logistic regression, xgboost, and decision tree models to predict the simulation outcome from these geometric features alone. To explore neural network ML techniques, we trained both a three - and a four-hidden layer multilayer perceptron feed forward neural networks (NN), again predicting simulation outcomes from these geometric features alone. ML and NN models were trained on 70% of randomly selected segments and the remaining 30% was used for validation for both baseline and augmented populations.

Results: Stimulation in the baseline population (30 patient models) resulted in reentry in 21.8% of sites tested; in the augmented population (129 total patient models) reentry occurred in 13.0% of sites tested. ML and NN models ranged in mean accuracy from 0.83 to 0.86 for the baseline population, improving to 0.88 to 0.89 in all cases.

Conclusion: Machine learning techniques, combined with patient-specific, image-based computational simulations, can provide key clinical insights with high accuracy rapidly and efficiently. In the case of sparse or missing patient data, simulation-supported data augmentation can be employed to further improve predictive results for patient benefit. This work paves the way for using data-driven simulations for prediction of dangerous arrhythmia in MI patients.

The Role of Artificial Intelligence and Machine Learning in Clinical Cardiac Electrophysiology

In recent years, artificial intelligence (AI) has found numerous applications in cardiology due in part to large digitized datasets and the evolution of high performance computing. In the discipline of cardiac electrophysiology (EP), a number of clinical, imaging, and electrical waveform data are considered in the diagnosis, prognostication and management of arrhythmias, which lend themselves well to automation through AI. But equally relevant, AI offers a unique opportunity to discover novel EP concepts and improve clinical care through its inherent, hierarchical tenets of self-learning. This review will focus on the application of AI in clinical EP and summarize state-of-the art, large, clinical studies in the following key domains: (1) ECG-based arrhythmia and disease classification, (2) atrial fibrillation source detection, (3) substrate and risk assessment for atrial fibrillation and ventricular tachyarrhythmias, and (4) predicting outcomes after cardiac resynchronization therapy. Many are small, single-center, proof-of-concept investigations, but they still demonstrate groundbreaking performance of deep learning, a subdomain of AI, which surpasses traditional statistical analysis. Larger studies, for instance classifying arrhythmias from ECG recordings, have further provided external validation of their high accuracy. Ultimately, the performance of AI is dependent on the quality of the input data and the rigor of algorithm development. The field is still nascent and several barriers will need to be overcome, including prospective validation in large, well-labelled datasets and more seamless information technology-based data collection/integration, before AI can be adopted into broader clinical EP practice. This review will conclude with a discussion of these challenges and future work.