Skip to Main content Skip to Navigation
New interface
Theses

Image-based Personalised Models of Cardiac Electrophysiology for Ventricular Tachycardia Therapy Planning

Abstract : Acute infarct survival rates have drastically improved over the last decades, mechanically increasing chronic infarct related affections.Among these affections, ischaemic ventricular tachycardia (VT) is a particularly serious arrhythmia that can lead to the often lethal ventricular fibrillation. VT can be treated by radio frequency ablation of the arrhythmogenic substrate.The first phase of this long and risky interventional cardiology procedure is an electrophysiological (EP) exploration of the heart.This phase aims at localising the ablation targets, notably by inducing the arrhythmia in a controlled setting. In this work we propose to re-create this exploration phase in silico, by personalising cardiac EP models.We show that key information about infarct scar location and heterogeneity can be automatically obtained by a deep learning-based automated segmentation of the myocardium on computed tomography (CT) images.Our goal is to use this information to run patient-specific simulations of depolarisation wave propagation in the myocardium, mimicking the interventional cardiology exploration phase.We start by studying the relationship between the depolarisation wave propagation velocity and the left ventricular wall thickness to personalise an Eikonal model, an approach that can successfully reproduce periodic activation maps of the left ventricle recorded during VT.We then propose efficient algorithms to detect the repolarisation wave on unipolar electrograms (UEG), that we use to analyse the UEGs embedded in such intra-cardiac recordings.Thanks to a multimodal registration between these recordings and CT images, we establish relationships between action potential durations/restitution properties and left ventricular wall thickness.These relationships are finally used to parametrise a reaction-diffusion model able to reproduce interventional cardiologists' induction protocols that trigger realistic and documented VTs. inteinterventional cardiologists' induction protocols that trigger realistic and documented VTs.
Document type :
Theses
Complete list of metadata

https://hal.archives-ouvertes.fr/tel-03147908
Contributor : ABES STAR :  Contact
Submitted on : Tuesday, April 27, 2021 - 12:34:15 PM
Last modification on : Saturday, June 25, 2022 - 11:50:36 PM

File

2020COAZ4097.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03147908, version 2

Collections

Citation

Nicolas Cedilnik. Image-based Personalised Models of Cardiac Electrophysiology for Ventricular Tachycardia Therapy Planning. Medical Imaging. Université Côte d'Azur, 2020. English. ⟨NNT : 2020COAZ4097⟩. ⟨tel-03147908v2⟩

Share

Metrics

Record views

155

Files downloads

57