Localization of Origins of Premature Ventricular Contraction by Means of Convolutional Neural Network From 12-Lead ECG

Deep learning for electrocardiogram interpretation: Bench to bedside

BACKGROUND

Recent advancements in deep learning (DL), a subset of artificial intelligence, have shown the potential to automate and improve disease recognition, phenotyping and prediction of disease onset and outcomes by analysing various sources of medical data. The electrocardiogram (ECG) is a valuable tool for diagnosing and monitoring cardiovascular conditions.

METHODS

The implementation of DL in ECG analysis has been used to detect and predict rhythm abnormalities and conduction abnormalities, ischemic and structural heart diseases, with performance comparable to physicians. However, despite promising development of DL algorithms for automatic ECG analysis, the integration of DL-based ECG analysis and deployment of medical devices incorporating these algorithms into routine clinical practice remains limited.

RESULTS

This narrative review highlights the applications of DL in 12-lead ECG analysis. Furthermore, we review randomized controlled trials that assess the clinical effectiveness of these DL tools. Finally, it addresses different key barriers to widespread implementation in clinical practice, including regulatory hurdles, algorithm transparency and data privacy concerns.

CONCLUSIONS

By outlining both the progress and the obstacles in this field, this review aims to provide insights into how DL could shape the future of ECG analysis and enhance cardiovascular care in daily clinical practice.

Solving the Inverse Problem of Electrocardiography for Cardiac Digital Twins: A Survey

Cardiac digital twins (CDTs) are personalized virtual representations used to understand complex cardiac mechanisms. A critical component of CDT development is solving the ECG inverse problem, which enables the reconstruction of cardiac sources and the estimation of patient-specific electrophysiology (EP) parameters from surface ECG data. Despite challenges from complex cardiac anatomy, noisy ECG data, and the ill-posed nature of the inverse problem, recent advances in computational methods have greatly improved the accuracy and efficiency of ECG inverse inference, strengthening the fidelity of CDTs. This paper aims to provide a comprehensive review of the methods for solving ECG inverse problems, their validation strategies, their clinical applications, and their future perspectives. For the methodologies, we broadly classify state-of-the-art approaches into two categories: deterministic and probabilistic methods, including both conventional and deep learning-based techniques. Integrating physics laws with deep learning models holds promise, but challenges such as capturing dynamic electrophysiology accurately, accessing accurate domain knowledge, and quantifying prediction uncertainty persist. Integrating models into clinical workflows while ensuring interpretability and usability for healthcare professionals is essential. Overcoming these challenges will drive further research in CDTs.

Spectral Cross-Domain Neural Network With Soft-Adaptive Threshold Spectral Enhancement

Electrocardiography (ECG) signals can be considered as multivariable time series (TS). The state-of-the-art ECG data classification approaches, based on either feature engineering or deep learning techniques, treat separately spectral and time domains in machine learning systems. No spectral-time domain communication mechanism inside the classifier model can be found in current approaches, leading to difficulties in identifying complex ECG forms. In this article, we proposed a novel deep learning model named spectral cross-domain neural network (SCDNN) with a new block called soft-adaptive threshold spectral enhancement (SATSE), to simultaneously reveal the key information embedded in spectral and time domains inside the neural network. More precisely, the domain-cross information is captured by a general convolutional neural network (CNN) backbone, and different information sources are merged by a self-adaptive mechanism to mine the connection between time and spectral domains. In SATSE, the knowledge from time and spectral domains is extracted via the fast Fourier transformation (FFT) with soft trainable thresholds in modified sigmoid functions. The proposed SCDNN is tested with several classification tasks implemented on the public ECG databases PTB-XL and CPSC2018. SCDNN outperforms the state-of-the-art approaches with a low computational cost regarding a variety of metrics in all classification tasks on both databases, by finding appropriate domains from the infinite spectral mapping. The convergence of the trainable thresholds in the spectral domain is also numerically investigated in this article. The robust performance of SCDNN provides a new perspective to exploit knowledge across deep learning models from time and spectral domains. The code repository can be found: https://github.com/DL-WG/SCDNN-TS.

Machine Learning for Localization of Premature Ventricular Contraction Origins: A Review

Premature ventricular contraction (PVC) is one of the most common arrhythmias, originating from ectopic beats in the ventricles. Precision in localizing the origin of PVCs has long been a focal point in electrophysiology research. Machine learning (ML) has developed rapidly in the past two decades with increasingly widespread applications. With the increase of clinical data such as electrocardiograms (ECGs), computed tomography (CT), and magnetic resonance imaging (MRI), ML and its subfields, deep learning (DL), have become powerful analytical tools, playing an increasingly important role in electrophysiological research. In this review, we mainly provide an overview of the development of ML in the localization of PVC origins, including its applications, advantages, disadvantages, and future research directions. This information is intended to serve as a reference for clinicians and researchers, aiding them in better-utilizing ML techniques for the diagnosis and study of PVC origins.

BOATMAP: Bayesian Optimization Active Targeting for Monomorphic Arrhythmia Pace-mapping

Recent advances in machine learning and deep learning have presented new opportunities for learning to localize the origin of ventricular activation from 12-lead electrocardiograms (ECGs), an important step in guiding ablation therapies for ventricular tachycardia. Passively learning from population data is faced with challenges due to significant variations among subjects, and building a patient-specific model raises the open question of where to select pace-mapping data for training. This work introduces BOATMAP, a novel active learning approach designed to provide clinicians with interpretable guidance that progressively assists in locating the origin of ventricular activation from 12-lead ECGs. BOATMAP inverts the input-output relationship in traditional machine learning solutions to this problem and learns the similarity between a target ECG and a paced ECG as a function of the pacing site coordinates. Using Gaussian processes (GP) as a surrogate model, BOATMAP iteratively refines the estimated similarity landscape while providing suggestions to clinicians regarding the next optimal pacing site. Furthermore, it can incorporate constraints to avoid suggesting pacing in non-viable regions such as the core of the myocardial scar. Tested in a realistic simulation environment in various heart geometries and tissue properties, BOATMAP demonstrated the ability to accurately localize the origin of activation, achieving an average localization accuracy of 3.9±3.6mm with only 8.0±4.0 pacing sites. BOATMAP offers real-time interpretable guidance for accurate localization and enhancing clinical decision-making.

Role of Machine Learning and Artificial Intelligence in Arrhythmias and Electrophysiology

Machine learning (ML), a subset of artificial intelligence (AI) centered on machines learning from extensive datasets, stands at the forefront of a technological revolution shaping various facets of society. Cardiovascular medicine has emerged as a key domain for ML applications, with considerable efforts to integrate these innovations into routine clinical practice. Within cardiac electrophysiology, ML applications, especially in the automated interpretation of electrocardiograms, have garnered substantial attention in existing literature. However, less recognized are the diverse applications of ML in cardiac electrophysiology and arrhythmias, spanning basic science research on arrhythmia mechanisms, both experimental and computational, as well as contributions to enhanced techniques for mapping cardiac electrical function and translational research related to arrhythmia management. This comprehensive review delves into various ML applications within the scope of this journal, organized into 3 parts. The first section provides a fundamental understanding of general ML principles and methodologies, serving as a foundational resource for readers interested in exploring ML applications in arrhythmia research. The second part offers an in-depth review of studies in arrhythmia and electrophysiology that leverage ML methodologies, showcasing the broad potential of ML approaches. Each subject is thoroughly outlined, accompanied by a review of notable ML research advancements. Finally, the review delves into the primary challenges and future perspectives surrounding ML-driven cardiac electrophysiology and arrhythmias research.

A transferable in-silico augmented ischemic model for virtual myocardial perfusion imaging and myocardial infarction detection

This paper proposes an innovative approach to generate a generalized myocardial ischemia database by modeling the virtual electrophysiology of the heart and the 12-lead electrocardiography projected by the in-silico model can serve as a ready-to-use database for automatic myocardial infarction/ischemia (MI) localization and classification. Although the virtual heart can be created by an established technique combining the cell model with personalized heart geometry to observe the spatial propagation of depolarization and repolarization waves, we developed a strategy based on the clinical pathophysiology of MI to generate a heterogeneous database with a generic heart while maintaining clinical relevance and reduced computational complexity. First, the virtual heart is simplified into 11 regions that match the types and locations, which can be diagnosed by 12-lead ECG; the major arteries were divided into 3-5 segments from the upstream to the downstream based on the general anatomy. Second, the stenosis or infarction of the major or minor coronary artery branches can cause different perfusion drops and infarct sizes. We simulated the ischemic sites in different branches of the arteries by meandering the infarction location to elaborate on possible ECG representations, which alters the infraction's size and changes the transmembrane potential (TMP) of the myocytes associated with different levels of perfusion drop. A total of 8190 different case combinations of cardiac potentials with ischemia and MI were simulated, and the corresponding ECGs were generated by forward calculations. Finally, we trained and validated our in-silico database with a sparse representation classification (SRC) and tested the transferability of the model on the real-world Physikalisch Technische Bundesanstalt (PTB) database. The overall accuracies for localizing the MI region on the PTB data achieved 0.86, which is only 2% drop compared to that derived from the simulated database (0.88). In summary, we have shown a proof-of-concept for transferring an in-silico model to real-world database to compensate for insufficient data.

A "two-step classification" machine learning method for non-invasive localization of premature ventricular contraction origins based on 12-lead ECG

BACKGROUND

Premature ventricular contraction (PVC) is a type of cardiac arrhythmia that originates from ectopic pacemaker in the ventricles. The localization of the origin of PVC is essential for successful catheter ablation. However, most studies on non-invasive PVC localization focus on elaborate localization in specific regions of the ventricle. This study aims to propose a machine learning algorithm based on 12-lead electrocardiogram (ECG) data that can improve the accuracy of PVC localization in the whole ventricle.

METHODS

We collected 12-lead ECG data from 249 patients with spontaneous or pacing-induced PVCs. The ventricle was divided into 11 segments. In this paper, we propose a machine learning method consisting of two consecutive classification steps. In the first classification step, each PVC beat was labeled to one of the 11 ventricular segments using six features, including a newly proposed morphological feature called "Peak_index." Four machine learning methods were tested for comparative multi-classification performance and the best classifier result was kept to the next step. In the second classification step, a binary classifier was trained using a smaller combination of features to further differentiate segments that are easily confused.

RESULTS

The Peak_index as a proposed new classification feature combined with other features is suitable for whole ventricle classification by machine learning methods. The test accuracy of the first classification reached 75.87%. It is shown that a second classification for confusable categories can improve the classification results. After the second classification, the test accuracy reached 76.84%, and when a sample classified into adjacent segments was considered correct, the test "rank accuracy" was improved to 93.49%. The binary classification corrected 10% of the confused samples.

CONCLUSION

This paper proposes a "two-step classification" method to localize the origin of PVC beats into the 11 regions of the ventricle using non-invasive 12-lead ECG. It is expected to be a promising technique to be used in clinical settings to help guide ablation procedures.

Non-invasive localization of the ventricular excitation origin without patient-specific geometries using deep learning

Cardiovascular diseases account for 17 million deaths per year worldwide. Of these, 25% are categorized as sudden cardiac death, which can be related to ventricular tachycardia (VT). This type of arrhythmia can be caused by focal activation sources outside the sinus node. Catheter ablation of these foci is a curative treatment in order to inactivate the abnormal triggering activity. However, the localization procedure is usually time-consuming and requires an invasive procedure in the catheter lab. To facilitate and expedite the treatment, we present two novel localization support techniques based on convolutional neural networks (CNNs) that address these clinical needs. In contrast to existing methods, our approaches were designed to be independent of the patient-specific geometry and directly applicable to surface ECG signals, while also delivering a binary transmural position. Moreover, one of the method's outputs can be interpreted as several ranked solutions. The CNNs were trained on a dataset containing only simulated data and evaluated both on simulated test data and clinical data. On a novel large and open simulated dataset, the median test error was below 3 mm. The median localization error on the unseen clinical data ranged from 32 mm to 41 mm without optimizing the pre-processing and CNN to the clinical data. Interpreting the output of one of the approaches as ranked solutions, the best median error of the top-3 solutions decreased to 20 mm on the clinical data. The transmural position was correctly detected in up to 82% of all clinical cases. These results demonstrate a proof of principle to utilize CNNs to localize the activation source without the intrinsic need for patient-specific geometrical information. Furthermore, providing multiple solutions can assist physicians in identifying the true activation source amongst more than one possible location. With further optimization to clinical data, these methods have high potential to accelerate clinical interventions, replace certain steps within these procedures and consequently reduce procedural risk and improve VT patient outcomes.

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.

ECGNet: An Efficient Network for Detecting Premature Ventricular Complexes Based on ECG Images

BACKGROUND

Preoperative prediction of the origin site of premature ventricular complexes (PVCs) is critical for the success of operations. However, current methods are not efficient or accurate enough. In addition, among the proposed strategies, there are few good prediction methods for electrocardiogram (ECG) images combined with deep learning aspects.

METHODS

We propose ECGNet, a new neural network for the classification of 12-lead ECG images. In ECGNet, 609 ECG images from 310 patients who had undergone successful surgery in the Division of Cardiology, the First Affiliated Hospital of Soochow University, are utilized to construct the dataset. We adopt dense blocks, special convolution kernels and divergent paths to improve the performance of ECGNet. In addition, a new loss function is designed to address the sample imbalance situation, whose cause is the uneven distribution of cases themselves, which often occurs in the medical field. We also conduct extensive experiments in terms of network prediction accuracy to compare ECGNet with other networks, such as ResNet and DarkNet.

RESULTS

Our ECGNet achieves extremely high prediction accuracy (91.74%) and efficiency with very small datasets. Our newly proposed loss function can solve the problem of sample imbalance during the training process.

CONCLUSION

The proposed ECGNet can quickly and accurately realize the multiclassification of PVCs after training with little data. Our network has the potential to be helpful to doctors with a preoperative diagnosis of PVCs. We will continue to collect similar cases and perfect our network structure to further improve the accuracy of our network's prediction.

Identification of Coronary Culprit Lesion in ST Elevation Myocardial Infarction by Using Deep Learning

OBJECTIVE

Early revascularization of the occluded coronary artery in patients with ST elevation myocardial infarction (STEMI) has been demonstrated to decrease mortality and morbidity. Currently, physicians rely on features of electrocardiograms (ECGs) to identify the most likely location of coronary arteries related to an infarct. We sought to predict these culprit arteries more accurately by using deep learning.

METHODS

A deep learning model with a convolutional neural network (CNN) that incorporated ECG signals was trained on 384 patients with STEMI who underwent primary percutaneous coronary intervention (PCI) at a medical center. The performances of various signal preprocessing methods (short-time Fourier transform [STFT] and continuous wavelet transform [CWT]) with different lengths of input ECG signals were compared. The sensitivity and specificity for predicting each infarct-related artery and the overall accuracy were evaluated.

RESULTS

ECG signal preprocessing with STFT achieved fair overall prediction accuracy (79.3%). The sensitivity and specificity for predicting the left anterior descending artery (LAD) as the culprit vessel were 85.7% and 88.4%, respectively. The sensitivity and specificity for predicting the left circumflex artery (LCX) were 37% and 99%, respectively, and the sensitivity and specificity for predicting the right coronary artery (RCA) were 88.4% and 82.4%, respectively. Using CWT (Morlet wavelet) for signal preprocessing resulted in better overall accuracy (83.7%) compared with STFT preprocessing. The sensitivity and specificity were 93.46% and 80.39% for LAD, 56% and 99.7% for LCX, and 85.9% and 92.9% for RCA, respectively.

CONCLUSION

Our study demonstrated that deep learning with a CNN could facilitate the identification of the culprit coronary artery in patients with STEMI. Preprocessing ECG signals with CWT was demonstrated to be superior to doing so with STFT.

Low Complexity Binarized 2D-CNN Classifier for Wearable Edge AI Devices

Wearable Artificial Intelligence-of-Things (AIoT) devices exhibit the need to be resource and energy-efficient. In this paper, we introduced a quantized multilayer perceptron (qMLP) for converting ECG signals to binary image, which can be combined with binary convolutional neural network (bCNN) for classification. We deploy our model into a low-power and low-resource field programmable gate array (FPGA) fabric. The model requires 5.8× lesser multiply and accumulate (MAC) operations than known wearable CNN models. Our model also achieves a classification accuracy of 98.5%, sensitivity of 85.4%, specificity of 99.5%, precision of 93.3%, and F1-score of 89.2%, along with dynamic power dissipation of 34.9 μW.

Training machine learning models with synthetic data improves the prediction of ventricular origin in outflow tract ventricular arrhythmias

In order to determine the site of origin (SOO) in outflow tract ventricular arrhythmias (OTVAs) before an ablation procedure, several algorithms based on manual identification of electrocardiogram (ECG) features, have been developed. However, the reported accuracy decreases when tested with different datasets. Machine learning algorithms can automatize the process and improve generalization, but their performance is hampered by the lack of large enough OTVA databases. We propose the use of detailed electrophysiological simulations of OTVAs to train a machine learning classification model to predict the ventricular origin of the SOO of ectopic beats. We generated a synthetic database of 12-lead ECGs (2,496 signals) by running multiple simulations from the most typical OTVA SOO in 16 patient-specific geometries. Two types of input data were considered in the classification, raw and feature ECG signals. From the simulated raw 12-lead ECG, we analyzed the contribution of each lead in the predictions, keeping the best ones for the training process. For feature-based analysis, we used entropy-based methods to rank the obtained features. A cross-validation process was included to evaluate the machine learning model. Following, two clinical OTVA databases from different hospitals, including ECGs from 365 patients, were used as test-sets to assess the generalization of the proposed approach. The results show that V2 was the best lead for classification. Prediction of the SOO in OTVA, using both raw signals or features for classification, presented high accuracy values (>0.96). Generalization of the network trained on simulated data was good for both patient datasets (accuracy of 0.86 and 0.84, respectively) and presented better values than using exclusively real ECGs for classification (accuracy of 0.84 and 0.76 for each dataset). The use of simulated ECG data for training machine learning-based classification algorithms is critical to obtain good SOO predictions in OTVA compared to real data alone. The fast implementation and generalization of the proposed methodology may contribute towards its application to a clinical routine.

An Energy Efficient ECG Ventricular Ectopic Beat Classifier Using Binarized CNN for Edge AI Devices

Wearable Artificial Intelligence-of-Things (AIoT) requires edge devices to be resource and energy-efficient. In this paper, we design and implement an efficient binary convolutional neural network (bCNN) algorithm utilizing function-merging and block-reuse techniques to classify between Ventricular and non-Ventricular Ectopic Beat images. We deploy our model into a low-resource low-power field programmable gate array (FPGA) fabric. Our model achieves a classification accuracy of 97.3%, sensitivity of 91.3%, specificity of 98.1%, precision of 86.7%, and F1-score of 88.9%, along with dynamic power dissipation of only 10.5-μW.

Big Data in electrophysiology

The quantity of data produced and captured in medicine today is unprecedented. Technological improvements and automation have expanded the traditional statistical methods and enabled the analysis of Big Data. This has permitted the discovery of new associations with a granularity that was previously hidden to human eyes. In the first part of this review, the authors would like to provide an overview of basic Machine Learning (ML) principles and techniques in order to better understand their application in recent publications about cardiac arrhythmias. In the second part, ML-enabled advances in disease detection and diagnosis, outcome prediction, and novel disease characterization in topics like electrocardiography, atrial fibrillation, ventricular arrhythmias, and cardiac devices are presented. Finally, the limitations and challenges of applying ML in clinical practice, such as validation, replication, generalizability, and regulatory issues, are discussed. More carefully designed studies and collaborations are needed for ML to become feasible, trustworthy, accurate, and reproducible and to reach its full potential for patient-oriented precision medicine.

Applications of Machine Learning in Ambulatory ECG

The ambulatory ECG (AECG) is an important diagnostic tool for many heart electrophysiology-related cases. AECG covers a wide spectrum of devices and applications. At the core of these devices and applications are the algorithms responsible for signal conditioning, ECG beat detection and classification, and event detections. Over the years, there has been huge progress for algorithm development and implementation thanks to great efforts by researchers, engineers, and physicians, alongside the rapid development of electronics and signal processing, especially machine learning (ML). The current efforts and progress in machine learning fields are unprecedented, and many of these ML algorithms have also been successfully applied to AECG applications. This review covers some key AECG applications of ML algorithms. However, instead of doing a general review of ML algorithms, we are focusing on the central tasks of AECG and discussing what ML can bring to solve the key challenges AECG is facing. The center tasks of AECG signal processing listed in the review include signal preprocessing, beat detection and classification, event detection, and event prediction. Each AECG device/system might have different portions and forms of those signal components depending on its application and the target, but these are the topics most relevant and of greatest concern to the people working in this area.

Cobiveco: Consistent biventricular coordinates for precise and intuitive description of position in the heart - with MATLAB implementation

Ventricular coordinates are widely used as a versatile tool for various applications that benefit from a description of local position within the heart. However, the practical usefulness of ventricular coordinates is determined by their ability to meet application-specific requirements. For regression-based estimation of biventricular position, for example, a symmetric definition of coordinate directions in both ventricles is important. For the transfer of data between different hearts as another use case, the consistency of coordinate values across different geometries is particularly relevant. To meet these requirements, we compare different approaches to compute coordinates and present Cobiveco, a symmetric, consistent and intuitive biventricular coordinate system that builds upon existing coordinate systems, but overcomes some of their limitations. A novel one-way transfer error is introduced to assess the consistency of the coordinates. Normalized distances along bijective trajectories between two boundaries were found to be superior to solutions of Laplace's equation for defining coordinate values, as they show better linearity in space. Evaluation of transfer and linearity errors on 36 patient geometries revealed a more than 4-fold improvement compared to a state-of-the-art method. Finally, we show two application examples underlining the relevance for cardiac data processing. Cobiveco MATLAB code is available under a permissive open-source license.

Machine Learning Identification of Pro-arrhythmic Structures in Cardiac Fibrosis

Cardiac fibrosis and other scarring of the heart, arising from conditions ranging from myocardial infarction to ageing, promotes dangerous arrhythmias by blocking the healthy propagation of cardiac excitation. Owing to the complexity of the dynamics of electrical signalling in the heart, however, the connection between different arrangements of blockage and various arrhythmic consequences remains poorly understood. Where a mechanism defies traditional understanding, machine learning can be invaluable for enabling accurate prediction of quantities of interest (measures of arrhythmic risk) in terms of predictor variables (such as the arrangement or pattern of obstructive scarring). In this study, we simulate the propagation of the action potential (AP) in tissue affected by fibrotic changes and hence detect sites that initiate re-entrant activation patterns. By separately considering multiple different stimulus regimes, we directly observe and quantify the sensitivity of re-entry formation to activation sequence in the fibrotic region. Then, by extracting the fibrotic structures around locations that both do and do not initiate re-entries, we use neural networks to determine to what extent re-entry initiation is predictable, and over what spatial scale conduction heterogeneities appear to act to produce this effect. We find that structural information within about 0.5 mm of a given point is sufficient to predict structures that initiate re-entry with more than 90% accuracy.

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.

Computer Modeling of the Heart for ECG Interpretation—A Review

Computer modeling of the electrophysiology of the heart has undergone significant progress. A healthy heart can be modeled starting from the ion channels via the spread of a depolarization wave on a realistic geometry of the human heart up to the potentials on the body surface and the ECG. Research is advancing regarding modeling diseases of the heart. This article reviews progress in calculating and analyzing the corresponding electrocardiogram (ECG) from simulated depolarization and repolarization waves. First, we describe modeling of the P-wave, the QRS complex and the T-wave of a healthy heart. Then, both the modeling and the corresponding ECGs of several important diseases and arrhythmias are delineated: ischemia and infarction, ectopic beats and extrasystoles, ventricular tachycardia, bundle branch blocks, atrial tachycardia, flutter and fibrillation, genetic diseases and channelopathies, imbalance of electrolytes and drug-induced changes. Finally, we outline the potential impact of computer modeling on ECG interpretation. Computer modeling can contribute to a better comprehension of the relation between features in the ECG and the underlying cardiac condition and disease. It can pave the way for a quantitative analysis of the ECG and can support the cardiologist in identifying events or non-invasively localizing diseased areas. Finally, it can deliver very large databases of reliably labeled ECGs as training data for machine learning.

Automated Localization of Focal Ventricular Tachycardia From Simulated Implanted Device Electrograms: A Combined Physics-AI Approach

Background: Focal ventricular tachycardia (VT) is a life-threating arrhythmia, responsible for high morbidity rates and sudden cardiac death (SCD). Radiofrequency ablation is the only curative therapy against incessant VT; however, its success is dependent on accurate localization of its source, which is highly invasive and time-consuming. Objective: The goal of our study is, as a proof of concept, to demonstrate the possibility of utilizing electrogram (EGM) recordings from cardiac implantable electronic devices (CIEDs). To achieve this, we utilize fast and accurate whole torso electrophysiological (EP) simulations in conjunction with convolutional neural networks (CNNs) to automate the localization of focal VTs using simulated EGMs. Materials and Methods: A highly detailed 3D torso model was used to simulate ∼4000 focal VTs, evenly distributed across the left ventricle (LV), utilizing a rapid reaction-eikonal environment. Solutions were subsequently combined with lead field computations on the torso to derive accurate electrocardiograms (ECGs) and EGM traces, which were used as inputs to CNNs to localize focal sources. We compared the localization performance of a previously developed CNN architecture (Cartesian probability-based) with our novel CNN algorithm utilizing universal ventricular coordinates (UVCs). Results: Implanted device EGMs successfully localized VT sources with localization error (8.74 mm) comparable to ECG-based localization (6.69 mm). Our novel UVC CNN architecture outperformed the existing Cartesian probability-based algorithm (errors = 4.06 mm and 8.07 mm for ECGs and EGMs, respectively). Overall, localization was relatively insensitive to noise and changes in body compositions; however, displacements in ECG electrodes and CIED leads caused performance to decrease (errors 16-25 mm). Conclusion: EGM recordings from implanted devices may be used to successfully, and robustly, localize focal VT sources, and aid ablation planning.

12-Lead ECG arrhythmia classification using cascaded convolutional neural network and expert feature

Owing to widely available digital ECG data and recent advances in deep learning techniques, automatic ECG arrhythmia classification based on deep neural network has gained growing attention. However, existing neural networks are mainly validated on single‑lead ECG, not involving the correlation and difference between multiple leads, while multiple leads ECG provides more complete description of the cardiac activity in different directions. This paper proposes a 12‑lead ECG arrhythmia classification method using a cascaded convolutional neural network (CCNN) and expert features. The one-dimensional (1-D) CNN is firstly designed to extract features from each single‑lead signal. Subsequently, considering the temporal correlation and spatial variability between multiple leads, features are cascaded as input to two-dimensional (2-D) densely connected ResNet blocks to classify the arrhythmia. Furthermore, features based on expert knowledge are extracted and a random forest is applied to get a classification probability. Results from both CCNN and expert features are combined using the stacking technique as the final classification result. The method has been validated against the first China ECG Intelligence Challenge, obtaining a final score of 86.5% for classifying 12‑lead ECG data with multiple labels into 9 categories.

A Novel Model Based on Spatial and Morphological Domains to Predict the Origin of Premature Ventricular Contraction

Pace mapping is commonly used to locate the origin of ventricular arrhythmias, especially premature ventricular contraction (PVC). However, this technique relies on clinicians’ ability to rapidly interpret ECG data. To avoid time-consuming interpretation of ECG morphology, some automated algorithms or computational models have been explored to guide the ablation. Inspired by these studies, we propose a novel model based on spatial and morphological domains. The purpose of this study is to assess this model and compare it with three existing models. The data are available from the Experimental Data and Geometric Analysis Repository database in which three in vivo PVC patients are included. To measure the hit rate (A hit occurs when the predicted site is within 15 mm of the target) of different algorithms, 47 target sites are tested. Moreover, to evaluate the efficiency of different models in narrowing down the target range, 54 targets are verified. As a result, the proposed algorithm achieves the most hits (37/47) and fewest misses (9/47), and it narrows down the target range most, from 27.62 ± 3.47 mm to 10.72 ± 9.58 mm among 54 target sites. It is expected to be applied in the real-time prediction of the origin of ventricular activation to guide the clinician toward the target site.

Machine Learning in Arrhythmia and Electrophysiology

Machine learning (ML), a branch of artificial intelligence, where machines learn from big data, is at the crest of a technological wave of change sweeping society. Cardiovascular medicine is at the forefront of many ML applications, and there is a significant effort to bring them into mainstream clinical practice. In the field of cardiac electrophysiology, ML applications have also seen a rapid growth and popularity, particularly the use of ML in the automatic interpretation of ECGs, which has been extensively covered in the literature. Much lesser known are the other aspects of ML application in cardiac electrophysiology and arrhythmias, such as those in basic science research on arrhythmia mechanisms, both experimental and computational; in the development of better techniques for mapping of cardiac electrical function; and in translational research related to arrhythmia management. In the current review, we examine comprehensively such ML applications as they match the scope of this journal. The current review is organized in 3 parts. The first provides an overview of general ML principles and methodologies that will afford readers of the necessary information on the subject, serving as the foundation for inviting further ML applications in arrhythmia research. The basic information we provide can serve as a guide on how one might design and conduct an ML study. The second part is a review of arrhythmia and electrophysiology studies in which ML has been utilized, highlighting the broad potential of ML approaches. For each subject, we outline comprehensively the general topics, while reviewing some of the research advances utilizing ML under the subject. Finally, we discuss the main challenges and the perspectives for ML-driven cardiac electrophysiology and arrhythmia research.

Real-Time Remote Health Monitoring System Driven by 5G MEC-IoT

Telemedicine over Internet of Things (IoT) generates an unprecedented amount of data, which further requires transmission, analysis, and storage. Deploying cloud computing to handle data of this magnitude will introduce unacceptable data analysis latency and high storage costs. Thus, mobile edge computing (MEC) deployed between the cloud and users, which is close to the nodes of data generation, can tackle these problems in 5G scenarios with the help of artificial intelligence. This paper proposes a telemedicine system based on MEC and artificial intelligence for remote health monitoring and automatic disease diagnosis. The integration of different technologies such as computers, medicine, and telecommunications will significantly improve the efficiency of patient treatment and reduce the cost of health care.

Automatic classification of healthy and disease conditions from images or digital standard 12-lead electrocardiograms

Standard 12-lead electrocardiography (ECG) is used as the primary clinical tool to diagnose changes in heart function. The value of automated 12-lead ECG diagnostic approaches lies in their ability to screen the general population and to provide a second opinion for doctors. Yet, the clinical utility of automated ECG interpretations remains limited. We introduce a two-way approach to an automated cardiac disease identification system using standard digital or image 12-lead ECG recordings. Two different network architectures, one trained using digital signals (CNN-dig) and one trained using images (CNN-ima), were generated. An open-source dataset of 41,830 classified standard ECG recordings from patients and volunteers was generated. CNN-ima was trained to identify atrial fibrillation (AF) using 12-lead ECG digital signals and images that were also transformed to mimic mobile device camera-acquired ECG plot snapshots. CNN-dig accurately (92.9–100%) identified every possible combination of the eight most-common cardiac conditions. Both CNN-dig and CNN-ima accurately (98%) detected AF from standard 12-lead ECG digital signals and images, respectively. Similar classification accuracy was achieved with images containing smartphone camera acquisition artifacts. Automated detection of cardiac conditions in standard digital or image 12-lead ECG signals is feasible and may improve current diagnostic methods.

A hybrid machine learning approach to localizing the origin of ventricular tachycardia using 12-lead electrocardiograms

BACKGROUND

Machine learning models may help localize the site of origin of ventricular tachycardia (VT) using 12-lead electrocardiograms. However, population-based models suffer from inter-subject anatomical variations within ECG data, while patient-specific models face the open challenge of what pacing data to collect for training.

METHODS

This study presents and validates the first hybrid model that combines population and patient-specific machine learning for rapid "computer-guided pace-mapping". A population-based deep learning model was first trained offline to disentangle inter-subject variations and regionalize the site of VT origin. Given a new patient with a target VT, an on-line patient-specific model -- after being initialized by the population-based prediction -- was then built in real time by actively suggesting where to pace next and improving the prediction with each added pacing data, progressively guiding pace-mapping towards the site of VT origin.

RESULTS

The population model was trained on pace-mapping data from 38 patients and the patient-specific model was subsequently tuned on one patient. The resulting hybrid model was tested on a separate cohort of eight patients in localizing 1) 193 LV endocardial pacing sites, and 2) nine VTs with clinically determined exit sites. The hybrid model achieved a localization error of 5.3 ± 2.6 mm using 5.4 ± 2.5 pacing sites in localizing LV pacing sites, achieving a significantly higher accuracy with a significantly smaller amount of training sites in comparison to models without active guidance.

CONCLUSION

The presented hybrid model has the potential to assist rapid pace-mapping of interventional targets in VT.

In silico pace-mapping: prediction of left vs. right outflow tract origin in idiopathic ventricular arrhythmias with patient-specific electrophysiological simulations

AIMS

A pre-operative non-invasive identification of the site of origin (SOO) of outflow tract ventricular arrhythmias (OTVAs) is important to properly plan radiofrequency ablation procedures. Although some algorithms based on electrocardiograms (ECGs) have been developed to predict left vs. right ventricular origins, their accuracy is still limited, especially in complex anatomies. The aim of this work is to use patient-specific electrophysiological simulations of the heart to predict the SOO in OTVA patients.

METHODS AND RESULTS

An in silico pace-mapping procedure was designed and used on 11 heart geometries, generating for each case simulated ECGs from 12 clinically plausible SOO. Subsequently, the simulated ECGs were compared with patient ECG data obtained during the clinical tachycardia using the 12-lead correlation coefficient (12-lead ρ). Left ventricle (LV) vs. right ventricle (RV) SOO was estimated by computing the LV/RV ratio for each patient, obtained by dividing the average 12-lead ρ value of the LV- and RV-SOO simulated ECGs, respectively. Simulated ECGs that had virtual sites close to the ablation points that stopped the arrhythmia presented higher correlation coefficients. The LV/RV ratio correctly predicted LV vs. RV SOO in 10/11 cases; 1.07 vs. 0.93 P < 0.05 for 12-lead ρ.

CONCLUSION

The obtained results demonstrate the potential of the developed in silico pace-mapping technique to complement standard ECG for the pre-operative planning of complex ventricular arrhythmias.

Prospective Assessment of an Automated Intraprocedural 12-Lead ECG-Based System for Localization of Early Left Ventricular Activation

BACKGROUND

To facilitate ablation of ventricular tachycardia (VT), an automated localization system to identify the site of origin of left ventricular activation in real time using the 12-lead ECG was developed. The objective of this study was to prospectively assess its accuracy.

METHODS

The automated site of origin localization system consists of 3 steps: (1) localization of ventricular segment based on population templates, (2) population-based localization within a segment, and (3) patient-specific site localization. Localization error was assessed by the distance between the known reference site and the estimated site.

RESULTS

In 19 patients undergoing 21 catheter ablation procedures of scar-related VT, site of origin localization accuracy was estimated using 552 left ventricular endocardial pacing sites pooled together and 25 VT-exit sites identified by contact mapping. For the 25 VT-exit sites, localization error of the population-based localization steps was within 10 mm. Patient-specific site localization achieved accuracy of within 3.5 mm after including up to 11 pacing (training) sites. Using 3 remotes (67.8±17.0 mm from the reference VT-exit site), and then 5 close pacing sites, resulted in localization error of 7.2±4.1 mm for the 25 identified VT-exit sites. In 2 emulated clinical procedure with 2 induced VTs, the site of origin localization system achieved accuracy within 4 mm.

CONCLUSIONS

In this prospective validation study, the automated localization system achieved estimated accuracy within 10 mm and could thus provide clinical utility.

Localization of origins of premature ventricular contraction in the whole ventricle based on machine learning and automatic beat recognition from 12-lead ECG

OBJECTIVE

The localization of origins of premature ventricular contraction (PVC) is the key factor for the success of ablation of ventricular arrhythmias. Existing methods rely heavily on manual extraction of PVC beats, which limits their application to the automatic PVC recognition from long-term data recorded by ECG monitors before and during operation. In addition, research identifying PVC sources in the whole ventricle have not been reported. The purpose of this study was to validate the feasibility of localization of origins of PVC in the whole ventricle and to explore an automatic algorithm for recognition of PVC beats based on long-term 12-lead ECG.

APPROACH

This study included 249 patients with spontaneous PVCs or pacing-induced PVCs. A novel algorithm was used to automatically extract PVC beats from a massive amount of original ECG data, which was collected by different acquisition devices. After clustering and labelling, 374 sample groups, each containing dozens to hundreds of PVC beats, formed the entire dataset of 11 categories corresponding to 11 regions of PVC origins in the whole ventricle. To choose the best classification model for the current task, four machine learning methods, support vector machine (SVM), random forest (RF), gradient-boosting decision tree (GBDT) and Gaussian naïve Bayes (GNB), were compared by randomly selecting 70% of the entire dataset (sample groups = 257) for training and the remaining 30% (sample groups = 117) for testing. The average performance of each model was estimated by the bootstrap method using 1000 resampling trials.

MAIN RESULTS

For PVC beat recognition, the achieved testing accuracy, sensitivity and specificity is 97.6%, 98.3% and 96.7%, respectively. For localization purpose, the achieved testing accuracy varies slightly from 70.7% to 74.1% among four classifiers, and when neighboring regions were combined, the testing rank accuracy is improved to a range of 91.5% to 93.2%.

SIGNIFICANCE

The proposed algorithm can automatically recognize PVC beats and map them to one of the 11 regions in the whole ventricle. Owing to the high accuracy of PVC beat recognition and the capability to target the potential PVC origins in multi regions, it is expected to be a predominant technique being used in clinical settings to automatically analyze huge ECG data before and during operation so as to replace the tedious manual identification.

Reduced leadset selection and performance evaluation in the inverse problem of electrocardiography for reconstructing the ventricularly paced electrograms

OBJECTIVE

Noninvasive electrocardiographic imaging (ECGI) is used for obtaining high-resolution images of the electrical activity of the heart, and is a powerful method with the potential to detect certain arrhythmias. However, there is no 'best' lead configuration in the literature to measure the torso potentials. This paper evaluates ECGI reconstructions using various reduced leadset configurations, explores whether one can find a common reduced leadset configuration that can accurately reconstruct the electrograms for datasets with different pacing sites, and compares two activation time estimation methods.

APPROACH

We used 23 ventricularly-paced datasets with pacing sites on different regions of the epicardium. Starting with a full 192‑leadset, we found "optimized" reduced leadsets specific to each dataset; we considered 64‑lead and 32‑lead configurations. Based on the histogram of individual "optimized" lead selections, we found a common reduced leadset. We compared the ECGI reconstructions and activation times of the individually optimized lead configurations with the common lead configurations.

RESULTS

Both 64‑lead configurations had similar performances to the 192‑leadset. 32‑leadset configurations, on the other hand, yielded noisy reconstructions, which affected their performance.

SIGNIFICANCE

There are no statistically significant differences in the performance of the inverse solutions when a 64‑lead common reduced leadset is used to estimate the electrograms and their respective pacing sites compared to using the full leadset. 32‑lead configurations, on the other hand, require a more careful study to improve their performance. The activation time method used significantly affects the pacing site estimation performance, especially with fewer electrodes.

Application of Neural Networks to 12-Lead Electrocardiography - Current Status and Future Directions

The 12-lead electrocardiogram (ECG) is a fast, non-invasive, powerful tool to diagnose or to evaluate the risk of various cardiac diseases. The vast majority of arrhythmias are diagnosed solely on 12-lead ECG. Initial detection of myocardial ischemia such as myocardial infarction (MI), acute coronary syndrome (ACS) and effort angina is also dependent upon 12-lead ECG. ECG reflects the electrophysiological state of the heart through body mass, and thus contains important information on the electricity-dependent function of the human heart. Indeed, 12-lead ECG data are complex. Therefore, the clinical interpretation of 12-lead ECG requires intense training, but still is prone to interobserver variability. Even with rich clinically relevant data, non-trained physicians cannot efficiently use this powerful tool. Furthermore, recent studies have shown that 12-lead ECG may contain information that is not recognized even by well-trained experts but which can be extracted by computer. Artificial intelligence (AI) based on neural networks (NN) has emerged as a strong tool to extract valuable information from ECG for clinical decision making. This article reviews the current status of the application of NN-based AI to the interpretation of 12-lead ECG and also discusses the current problems and future directions.

Sequential Factorized Autoencoder for Localizing the Origin of Ventricular Activation From 12-Lead Electrocardiograms

OBJECTIVE

This work presents a novel approach to handle the inter-subject variations existing in the population analysis of ECG, applied for localizing the origin of ventricular tachycardia (VT) from 12-lead electrocardiograms (ECGs).

METHODS

The presented method involves a factor disentangling sequential autoencoder (f-SAE) - realized in both long short-term memory (LSTM) and gated recurrent unit (GRU) networks - to learn to disentangle the inter-subject variations from the factor relating to the location of origin of VT. To perform such disentanglement, a pair-wise contrastive loss is introduced.

RESULTS

The presented methods are evaluated on ECG dataset with 1012 distinct pacing sites collected from scar-related VT patients during routine pace-mapping procedures. Experiments demonstrate that, for classifying the origin of VT into the predefined segments, the presented f-SAE improves the classification accuracy by 8.94% from using prescribed QRS features, by 1.5% from the supervised deep CNN network, and 5.15% from the standard SAE without factor disentanglement. Similarly, when predicting the coordinates of the VT origin, the presented f-SAE improves the performance by 2.25 mm from using prescribed QRS features, by 1.18 mm from the supervised deep CNN network and 1.6 mm from the standard SAE.

CONCLUSION

These results demonstrate the importance as well as the feasibility of the presented f-SAE approach for separating inter-subject variations when using 12-lead ECG to localize the origin of VT.

SIGNIFICANCE

This work suggests the important research direction to deal with the well-known challenge posed by inter-subject variations during population analysis from ECG signals.

Localization of Ventricular Activation Origin from the 12-Lead ECG: A Comparison of Linear Regression with Non-Linear Methods of Machine Learning

We have previously developed an automated localization method based on multiple linear regression (MLR) model to estimate the activation origin on a generic left-ventricular (LV) endocardial surface in real time from the 12-lead ECG. The present study sought to investigate whether machine learning-namely, random-forest regression (RFR) and support-vector regression (SVR)-can improve the localization accuracy compared to MLR. For 38 patients the 12-lead ECG was acquired during LV endocardial pacing at 1012 sites with known coordinates exported from an electroanatomic mapping system; each pacing site was then registered to a generic LV endocardial surface subdivided into 16 segments tessellated into 238 triangles. ECGs were reduced to one variable per lead, consisting of 120-ms time integral of the QRS. To compare three regression models, the entire dataset ([Formula: see text]) was partitioned at random into a design set with 80% and a test set with the remaining 20% of the entire set, and the localization error-measured as geodesic distance on the generic LV surface-was assessed. Bootstrap method with replacement, using 1000 resampling trials, estimated each model's error distribution for the left-out sample ([Formula: see text]). In the design set ([Formula: see text]), the mean accuracy was 8.8, 12.1, and 12.9 mm, respectively for SVR, RVR and MLR. In the test set ([Formula: see text]), the mean value of the localization error in the SVR model was consistently lower than the other two models, both in comparison with the MLR (11.4 vs. 12.5 mm), and with the RFR (11.4 vs. 12.0 mm); the RFR model was also better than the MLR model for estimating localization accuracy (12.0 vs. 12.5 mm). The bootstrap method with 1,000 trials confirmed that the SVR and RFR models had significantly higher predictive accurate than the MLR in the bootstrap assessment with the left-out sample (SVR vs. MLR ([Formula: see text]), RFR vs. MLR ([Formula: see text])). The performance comparison of regression models showed that a modest improvement in localization accuracy can be achieved by SVR and RFR models, in comparison with MLR. The "population" coefficients generated by the optimized SVR model from our dataset are superior to the previously-derived "population" coefficients generated by the MLR model and can supersede them to improve the localization of ventricular activation on the generic LV endocardial surface.

Localization of Activation Origin on Patient-Specific Epicardial Surface by Empirical Bayesian Method

OBJECTIVE

Ablation treatment of ventricular arrhythmias can be facilitated by pre-procedure planning aided by electrocardiographic inverse solution, which can help to localize the origin of arrhythmia. Our aim was to improve localization accuracy of the inverse solution by using a novel Bayesian approach.

METHODS

The inverse problem of electrocardiography was solved by reconstructing epicardial potentials from 120 body-surface electrocardiograms and from patient-specific geometry of the heart and torso for four patients suffering from scar-related ventricular tachycardia who underwent epicardial catheter mapping, which included pace-mapping. Simulations using dipole sources in patient-specific geometry were also performed. The proposed method, using dynamic spatio-temporal a priori constraints of the solution, was compared with classical Tikhonov methods based on fixed constraints.

RESULTS

The mean localization error of the proposed method for all available pacing sites (n=78) was significantly smaller than that achieved by Tikhonov methods; specifically, the localization accuracy for pacing in the normal tissue (n=17) was [Formula: see text] mm (mean ± SD) versus [Formula: see text] mm reported in the previous study using the same clinical data and Tikhonov regularization. Simulation experiments further supported these clinical findings.

CONCLUSION

The promising results of in vivo and in silico experiments presented in this study provide a strong incentive to pursuing further investigation of data-driven Bayesian methods in solving the electrocardiographic inverse problem.

SIGNIFICANCE

The proposed approach to localizing origin of ventricular activation sequence may have important applications in pre-procedure assessment of arrhythmias and in guiding their ablation treatment.

Deep Learning on 1-D Biosignals: a Taxonomy-based Survey

OBJECTIVES

 Deep learning models such as convolutional neural networks (CNNs) have been applied successfully to medical imaging, but biomedical signal analysis has yet to fully benefit from this novel approach. Our survey aims at (i) reviewing deep learning techniques for biosignal analysis in computer- aided diagnosis; and (ii) deriving a taxonomy for organizing the growing number of applications in the field.

METHODS

 A comprehensive literature research was performed using PubMed, Scopus, and ACM. Deep learning models were classified with respect to the (i) origin, (ii) dimension, and (iii) type of the biosignal as input to the deep learning model; (iv) the goal of the application; (v) the size and (vi) type of ground truth data; (vii) the type and (viii) schedule of learning the network; and (ix) the topology of the model.

RESULTS

 Between January 2010 and December 2017, a total 71 papers were published on the topic. The majority (n = 36) of papers are on electrocariography (ECG) signals. Most applications (n = 25) aim at detection of patterns, while only a few (n = 6) at predection of events. Out of 36 ECG-based works, many (n = 17) relate to multi-lead ECG. Other biosignals that have been identified in the survey are electromyography, phonocardiography, photoplethysmography, electrooculography, continuous glucose monitoring, acoustic respiratory signal, blood pressure, and electrodermal activity signal, while ballistocardiography or seismocardiography have yet to be analyzed using deep learning techniques. In supervised and unsupervised applications, CNNs and restricted Boltzmann machines are the most and least frequently used, (n = 34) and (n = 15), respectively.

CONCLUSION

 Our key-code classification of relevant papers was used to cluster the approaches that have been published to date and demonstrated a large variability of research with respect to data, application, and network topology. Future research is expected to focus on the standardization of deep learning architectures and on the optimization of the network parameters to increase performance and robustness. Furthermore, application-driven approaches and updated training data from mobile recordings are needed.

Transfer Learning From Simulations on a Reference Anatomy for ECGI in Personalized Cardiac Resynchronization Therapy

GOAL

Noninvasive cardiac electrophysiology (EP) model personalisation has raised interest for instance in the scope of predicting EP cardiac resynchronization therapy (CRT) response. However, the restricted clinical applicability of current methods is due in particular to the limitation to simple situations and the important computational cost.

METHODS

We propose in this manuscript an approach to tackle these two issues. First, we analyze more complex propagation patterns (multiple onsets and scar tissue) using relevance vector regression and shape dimensionality reduction on a large simulated database. Second, this learning is performed offline on a reference anatomy and transferred onto patient-specific anatomies in order to achieve fast personalized predictions online.

RESULTS

We evaluated our method on a dataset composed of 20 dyssynchrony patients with a total of 120 different cardiac cycles. The comparison with a commercially available electrocardiographic imaging (ECGI) method shows a good identification of the cardiac activation pattern. From the cardiac parameters estimated in sinus rhythm, we predicted five different paced patterns for each patient. The comparison with the body surface potential mappings (BSPM) measured during pacing and the ECGI method indicates a good predictive power.

CONCLUSION

We showed that learning offline from a large simulated database on a reference anatomy was able to capture the main cardiac EP characteristics from noninvasive measurements for fast patient-specific predictions.

SIGNIFICANCE

The fast CRT pacing predictions are a step forward to a noninvasive CRT patient selection and therapy optimisation, to help clinicians in these difficult tasks.