PERSONALISED MODELLING OF VIRCHOW'S TRIAD FOR STROKE RISK ASSESSMENT IN ATRIAL FIBRILLATION

点击翻译按钮获取中文摘要

Atrial Fibrillation (AF) is the most prevalent cardiac arrhythmia in clinical settings and is brought about by a series of uncoordinated electrical activations which diminish the mechanical function of the left atrium (LA). Consequently, AF is reported to increase the risk of stroke five-fold which may be attributed to increased blood stasis, endothelial damage and hypercoagubility in the LA, all of which are physiological factors thought to promote thrombosis in Virchow's Triad. Currently, stroke risk stratification for AF patients is performed using the CHA2DS2-VASc scores and is based on empirical population information. Oral anticoagulants (OACs) or novel OACs are prescribed to patients with moderate-to-high risk of stroke. However, these risk scores provide little information to assess if low-risk patients, where OACs are not given, are truly low-risk, given that termination of AF by catheter ablation may result in impaired contractility and may promote blood stasis. This investigation aims to develop and use mechanistic computational fluid dynamic models to help determine the propensity of clot formation in patients who are classified as low risk using current risk stratification indices.To achieve this goal, I will develop a workflow for performing 3D blood flow simulations on patient-specific LA geometries. 3D models will be reconstructed from time-varying Cine MRI data with motion-tracking applied through MIRTK routines. All flow simulations will be performed using our in-house finite-element based biomechanics simulator, CHeart, with mathematical modelling of blood coagulation implemented through reaction-diffusion-convection equations. CHeart will be used to simulate small-scale biochemical reactions of blood coagulation alongside large-scale 3D fluid dynamics in the LA. This will allow for personalised predictions of the likelihood of thrombus formation which can supplement current risk scores. Findings to be validated using 4D Flow MRI.