Electrophysiological remodelling in patients at elevated risk of ventricular arrhythmia and sudden cardiac death
Our research has two aims. Firstly, to study the electrical activity of the heart in individuals that have a higher risk of heart rhythm disturbance, called 'arrhythmia'. Secondly, to establish if we can develop methods to accurately predict which individuals are at increased risk of arrhythmia using existing demographic and clinical information. We know that medical problems associated with heart disease, such as obesity and diabetes, are becoming increasingly common and can cause arrhythmia. In some instances, these can be life-threatening. Our ability to identify patients at risk of arrhythmia has been limited to standard 12-lead electrical recordings (ECG). Using the data from the UK BioBank, we aim to conduct more detailed investigation to establish links between clinical measurements and the heart's structure and electrical activity. Using a method called machine learning, we also aim to develop ways to predict patients' current and/or future risk of arrhythmia. We will do this alongside another ongoing study, in which we will use a more sophisticated 252-lead ECG and combine it with heart scans (also called ECGI), to better understand how common medical conditions affect the heart's electrical activity. This study will therefore improve our understanding of how demographic and clinical factors affect or influence the electrical activity of the heart and guide future research.
Hypothesis: cardiometabolic phenotypes are associated with electrocardiographic and structural changes in the myocardium that increase risk of ventricular arrhythmia
Aims: To identify and characterise any correlations that exist between anthropometric data and proarrhythmic electrocardiographic features To identify and characterise any correlations that exist between anthropometric data and structural heart disease (cardiac MRI) To identify if and how morbidity of various aetiologies, particularly cardiometabolic and inflammatory conditions, alter electrocardiographic and structural (cardiac MRI) parameters that might predispose to ventricular arrhythmia To generate machine learning models to predict risk of arrhythmia that can translate into better clinical risk stratification
New scope: Additionally we aim to identify electrical (based on ECG) and structural (based on cardiac MRI parameters) phenotypes that predict cardiovascular outcomes, including all arrhythmias (e.g. atrial fibrillation, conduction disease, as well as ventricular arrhythmias), major adverse cardiovascular events and death. We aim to correlate electrical and structural phenotypes with demographics, biomarkers and comorbidity data to gain a deeper understanding of phenotype determinants and clinical significant. We aim to utilise our findings for detailed and personalised risk stratification and disease management. Through the use of machine learning we aim to identify novel features to define phenotypes and their clinical significance