Rethinking multiscale cardiac electrophysiology with machine learning and predictive modelling

Abstract

We review some of the latest approaches to analysing cardiac electrophysiology data using machine learning and predictive modelling. Cardiac arrhythmias, particularly atrial fibrillation, are a major global healthcare challenge. Treatment is often through catheter ablation, which involves the targeted localised destruction of regions of the myocardium responsible for initiating or perpetuating the arrhythmia. Ablation targets are either anatomically defined, or identified based on their functional properties as determined through the analysis of contact intracardiac electrograms acquired with increasing spatial density by modern electroanatomic mapping systems. While numerous quantitative approaches have been investigated over the past decades for identifying these critical curative sites, few have provided a reliable and reproducible advance in success rates. Machine learning techniques, including recent deep-learning approaches, offer a potential route to gaining new insight from this wealth of highly complex spatio-temporal information that existing methods struggle to analyse. Coupled with predictive modelling, these techniques offer exciting opportunities to advance the field and produce more accurate diagnoses and robust personalised treatment. We outline some of these methods and illustrate their use in making predictions from the contact electrogram and augmenting predictive modelling tools, both by more rapidly predicting future states of the system and by inferring the parameters of these models from experimental observations.

Automatic diagnosis of the 12-lead ECG using a deep neural network

“…Recurrent networks are used in [26,15]. A review of recent machine learning techniques applied for ECG automatic diagnosis is given in [27]. The aforementioned methods and others (such as random forest and bayesian methods) are compared and a more extensive list of references using those methods is provided.…”

Development of a pro-arrhythmic ex vivo intact human and porcine model: cardiac electrophysiological changes associated with cellular uncoupling

“…We developed signal processing algorithms to automatically extract characteristics of the electrogram morphology for every activation in the dataset in a consistent manner. The algorithm was adapted for the intact heart from the one we previously developed for the analysis of myocardial cell monolayers [3]. A broad range of EGM signal time-domain features were extracted for each individual recording.…”

“…A broad range of EGM signal time-domain features were extracted for each individual recording. The 19 features included were selected due to their known and established importance in cardiac electrophysiology from EGM and ECG interpretations, or as a region of interest previously investigated with myocardial cell monolayers that may provide further mechanistic insight [3]. To validate the accuracy of the automated interpretation, randomly selected EGMs were manually interpreted and compared with the feature extraction values.…”

“…Identifying the features with most significant changes from BL in these EGMs could be used to establish collectively which combination of features differentiate between normal and different specific pathological EGMs, leading to a greater understanding of how pathological myocardial electrophysiological function manifests in the clinically accessible EGM. To leverage the analysis of the EGMs, there is potential to apply machine learning techniques to automatically distinguish more subtle changes in EGM morphology between a variety of underlying pathologies [3]. Due to the ex vivo preparation's similarities to patients, with regard to size and baseline electrophysiological properties, the machine learning model could be applied to detect the same subtle changes in in vivo EGMs.…”

Challenges Associated with Interpreting Mechanisms of AF

“…registration,135 fibrotic remodelling118 and performing predictions on clinical timescales 136. Identifying the properties of the atrial substrate responsible for sustaining the arrhythmia (e.g.…”