Patient-specific computational fluid dynamics (CFD) modelling of the left ventricle (LV) is a promising technique for the visualisation of ventricular flow patterns throughout a cardiac cycle. While significant progress has been made in improving the physiological quality of such simulations, the methodologies involved for several key steps remain significantly operator-dependent to this day. This dependency limits both the efficiency of the process as well as the consistency of CFD results due to the labour-intensive nature of current methods as well as operator introduced uncertainties in the modelling process. In order to mitigate this dependency, we propose a semi-automated method for patient-specific computational flow modelling of the LV. Using magnetic resonance imaging derived coarse geometry data of a patient's LV endocardium shape throughout a cardiac cycle, we then proceed to refine the geometry to eliminate rough edges before reconstructing meshes for all time frames and finally numerically solving for the intra-ventricular flow. Using a sample of patient-specific volunteer data, we demonstrate that our semi-automated, minimal operator involvement approach is capable of yielding CFD results of the LV that are comparable to other clinically validated LV flow models in the literature.
Keywords: cardiac cycle; computational fluid dynamics; geometry reconstruction; left ventricles; patient-specific.