Project ID BE-MI2023_02


Co Supervisor 1A School of Biomedical Engineering & Imaging SciencesWebsite

Co Supervisor 1B School of Biomedical Engineering & Imaging SciencesWebsite

Predictive Modelling of Cardiovascular Diseases Using Physics-Informed Artificial Intelligence

Background: Cardiovascular diseases (CVD) are a leading cause of morbidity and mortality. Disruptions of electrophysiological mechanisms in the heart lead to cardiac arrhythmias that affectblood supply to the vital organs, and also greatly increase risks of thrombus formation and stroke. However, arrhythmogenic mechanisms and their effects on the blood floware poorly understood,and treatments are often deficient. Computational modelling emerged as a powerful toolfordissecting the mechanisms of arrhythmias and improving their treatments. However, high computational costs hinder application of the models.

Aims: The project will create fast and computationally-inexpensive tools for simulating complex CVD conditions,leadingto the development of a robust, reliable and rapid frameworkfor patient-specific cardiac simulations. This will be achieved through the development and application ofPhysics-informed Artificial Intelligence (AI). This advanced approach enablesintegration of the underlying laws of (Bio-) Physics into AI algorithms that predict outcomes in patients with CVD. The project will focus on two linked problems: 1) arrhythmiasin 3Datrial electrophysiology models and prediction of efficient treatments, and 2) blood stasis in 3D atrial blood flow models and prediction of associated risks of stroke.

Timeline: Year 1: Development of Physics-informed neural networks(PINN) for fast simulation of3D atrial models. Year 2: PINN application to simulate atrial arrhythmias and predict anti-arrhythmictreatments. Year 3: PINN application to simulate atrial blood flow and quantifyassociatedrisks of stroke. Skills.The student will gain advanced quantitative skills, including mathematical and biophysicalmodelling,AI andcomputer programming, andstrong interdisciplinary skillsin applying AI to solve CVD problems.

One representative publication from each co-supervisor:

[1] Muffoletto M, Qureshi A, Zeidan A, Muizniece L, Fu X, Zhao J, Roy A, Bates PA,Aslanidi O. (2021).Toward Patient-Specific Prediction of Ablation Strategies for Atrial Fibrillation Using Deep Learning. Frontiers in Physiology12:674106;

[2] Miller R, Kerfoot E, Mauger C, Ismail T, Young A,Nordsletten D. (2021). An Implementation of Patient-Specific Biventricular Mechanics Simulations With a Deep Learning and Computational Pipeline.Frontiers in Physiology12:716597;