Spatial Cardiac Dysfunction Assessment via Personalized Modelling from MRI

Cardiac MRI is a standard modality to diagnose cardiac abnormality and dysfunction as recommended by the expert group. Cardiac assessment in current clinic, however, is limited to visual inspection of MRI images with some manual measurements to quantify structural and functional defects before a diagnosis is established. In this study, we describe a novel 3D+time personalized LV modelling algorithm to facilitate clinical assessment of cardiac diseases from cine MRI, and we have targeted acute myocardial infarction (AMI) and hypertrophy cardiomyopathy (HCM). Spatial computation of wall thickness and wall dynamics in terms of maximum thickening across full cardiac cycle were performed on LV models of individual patient, before mapping to bull's eye diagram for regional analysis. The analysis was performed on 29 AMI and 5 HCM patients, and compared against 15 healthy subjects. Regional contractility dysfunction was presented and correlated to the presence of myocardial infarction and fibrosis as assessed through LGE MRI scans. Our approach provides 3D+time visualization of LV wall abnormality which is found useful to aid descriptive diagnosis of AMI and HCM, therefore suggestive of potential clinical utility.