Personalised 3D modelling of the right ventricle as an approach to ARVC

Lead Research Organisation: University of Oxford
Department Name: Computer Science

Abstract

1 Abstract
Arrhythmogenic Right Ventricular Cardiomyopathy (ARVC) is one of the leading causes of Sudden Cardiac Death in young individuals. Finer stratification and early diagnosis are vital for reducing the number of premature deaths caused by ARVC. This study aims to develop personalised high performance computer models based on electrocardiographic imaging for risk stratification in ARVC. Deeper insight ARVC on the human heart will be explored via coupled simulations of ionic and structural remodelling, by solving coupled systems of ODEs/PDEs, namely the mono and bi-domain models. Fractional diffusion models will also be implemented, as they have been proven an innovative way of approximating physiological dynamics.
2 Aims and Objectives
The main goal of the study is to provide mechanistic insight into the role of the underlying, pro-arrhythmic, electro-anatomical substrate in governing the initiation and maintenance of ventricular arrhythmias in ARVC. The ultimate goal is to propose novel treatments as well as improve the efficiency of existing ones(drugs, ablation and implantable defibrillators). In order to achieve this, we will create models incorporating coupled PDEs and stiff ODEs, based on human ventricular meshes of many millions of elements.The personalised computer models will provide a basic science and diagnostic tool for transforming the raw patient data into a 3D dynamical heart representations. Electro-cardiographic imagining (ECGi) data will be central to this study, as its high spatial resolution will be used to track complex dynamics of arrhythmias.
3 Novelty of the research
There is confusion in literature as to whether electrical imbalances predate structural disarrangements in the ventricles and what causes the onset of symptoms. In this study, we aim to provide more extensive, big data and modelling backed answers. Furthermore, we want to infer if particular forms of structural disarrangement are more harmful than others. The ECGi technique is a novel, non-invasive approach to monitoring the electrical activity of the heart, boasting very high spatial resolution (4-7mm) and quick recording of data. Only 12 hospitals worldwide have access to this technique, a fraction of which use it for research purposes. The remodelling aims to create a detailed, patient specific representation of the ventricles. Inter-cell variability will be accounted for by incorporating biological data. An accurate representation of the electrical dynamics in regions, plagued by micro disarrangements, requires a fine discretisation of the relevant PDEs. This highly detailed representation of the ventricular electrical dynamics requires vast computational power. The largest super-computer in the UK (ARCHER) will be used to perform the computation. Fractional diffusion models are another approach when characterising heterogeneous electrical signals, such as those in heart arrhythmias. In order to reduce possible inaccuracies in terms of model parameterasation of physical quantities, machine learning methods will be used on patient data in order to determine realistic values for the biological parameters.
4 Alignment with EPSRCs strategies
The project falls under the Health care technologies research scheme, more specifically to the Clinical Technologies sub unit. It is a modelling and basic science project, aimed to enhance understanding in areas of biomedical interest (Techniques for Biomedical Understanding) and in the design and development of technologies for advancing the detection and diagnosis of health conditions (Diagnostics).
5 Collaborators
A close research collaboration is established with prominent cardiologists at the Barts Heart Centre in London. They will provide the ECGi recording of a cohort of 100 patients suffering from various degrees of ARVC. In addition, they will inform of the clinical challenges and needs
when diagnosing and treating patients with ARVC.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/N509711/1 01/10/2016 30/09/2021
1746775 Studentship EP/N509711/1 03/10/2016 31/03/2020 Peter Marinov
 
Description I have managed to suggest a mechanism responsible for disrupting the electrical conduction system of the heart in patients suffering from ACM, formerly known as ARVC. The simulation experiments i have run confirm this mechanism is a valid one. This can potentially lead to new treatments and earlier diagnosis of this lethal disease. I have submitted a paper to a cardiology journal, which highlights my recent findings.
Exploitation Route Other people in the modelling community can use my tools to advance computational cardiovascular research.
Pharma companies could use my research to gain a better understand of what properties a cure targeting ACM should satisfy.
I will be liaising with Phillips (Netherlands) who want to incorporate our technology into their research tool pipeline.
Sectors Digital/Communication/Information Technologies (including Software),Healthcare,Pharmaceuticals and Medical Biotechnology

 
Description Phillips want to use our technology and incorporate it into their research and testing tools.
Sector Healthcare,Pharmaceuticals and Medical Biotechnology
Impact Types Economic

 
Title Pipeline for modelling fibrosis from MRI data 
Description The percentage of fibrosis data in a section of the ventricles is introduced into a 3D whole ventricular model of the patient. This is done by reducing the myocardial extracellular and intracellular conductivities in the Bidomain equation of electrical propagation. Hence, by incorporating this medical data, the models of the patient ventricles more closely mimic the patient condition and can explain why electrical conduction is disrupted in these patients. 
Type Of Material Model of mechanisms or symptoms - human 
Year Produced 2018 
Provided To Others? No  
Impact This tool will help to bring insight into the functioning of Arrhythmogenic Right Ventricular Cardiomyopathy as well as into the mechanisms which lead to lethal arrhythmias in this disease. 
 
Title Pipeline for Extraction of Ventricular Geometries from MRI data 
Description It is an extension/improvement of a pipeline to create detailed anatomical representations of heart-torso geometries from MRI data. This software is semi automatic and comes with a user manual. 
Type Of Technology Software 
Year Produced 2018 
Impact This software enabled me to create virtual mesh based representations of a cohort of 20 Arrhythmogenic Right Ventricular Cardiomyopathy patients and use them to solve finite element methods on them. I also did a geometric based analysis on the shapes of the ventricles of ARVC patients.