|In-Vivo Estimation of Physical Properties in Right Ventricular Mechanics|
|Supervisor: Professor Alejandro Frangi|
Right ventricular (RV) function may be impaired in a number of heart conditions, including myocardial infarction, congenital heart disease and cardiomyopathy. The function of the RV may also be affected in diseases of the left ventricle (LV) where it is difficult to ignore the complex nature of ventricular interaction. In the past, the importance of the LV in cardiac research has overshadowed the study of the RV. This neglect is, in part, because the RV is difficult to image. The wall of the RV myocardium is thin (2-5mm) when compared to that of the LV (7-11mm). Furthermore the RV has a complex geometry, eccentric motion and is heavily trabeculated, thus it does not offer the clearly defined endocardial margins typically seen in the LV.
Non-rigid image registration has been proposed as an automatic method for recovering 3D cardiac displacement fields from tagged Magnetic Resonance Imaging (tMRI) or from 3D echocardiographic image sequences. Such algorithms enable the extraction of a continuous and bijective 3D deformation field from which cardiac strains can be computed and, with additional information, myocardial stresses and elastic properties could be derived. Such properties might be related to the health status of the myocardium, be used as disease biomarkers, or used for personalisation of biomedical models of cardiac contraction. A number of extensions have been developed to incorporate biomedical contraints or to accelerate it using different computational schemes or GPU implementations.
The project aims at: a) extensive analysis of the literature to understand the anatomical and physiological properties of the RV; b) understanding of cardiac mechanics and, in particular, of RV mechanics; c) acquiring a database of RV magnetic resonance images in a group of control subjects (CS) and patients with pulmonary hypertension (PH); d) adapting the methods proposed by Modat et al. (2010) to the framework of temporal diffeomorphioc transformations; e) estimating RV myocardial strains in CS and HP subjects and analysing the difference.