RESUMO
This study aimed to create an imaging-derived patient-specific computational model of low-flow, low-gradient (LFLG) aortic stenosis (AS) to obtain biomechanics data about the left ventricle. LFLG AS is now a commonly recognized sub-type of aortic stenosis. There remains much controversy over its management, and investigation into ventricular biomechanics may elucidate pathophysiology and better identify patients for valve replacement. ECG-gated cardiac computed tomography images from a patient with LFLG AS were obtained to provide patient-specific geometry for the computational model. Surfaces of the left atrium, left ventricle (LV), and outflow track were segmented. A previously validated multi-scale, multi-physics computational human heart model was adapted to the patient-specific geometry, yielding a model consisting of 91,000 solid elements. This model was coupled to a virtual circulatory system and calibrated to clinically measured parameters from echocardiography and cardiac catheterization data. The simulation replicated key physiologic parameters within 10% of their clinically measured values. Global LV systolic myocardial stress was 7.1 ± 1.8 kPa. Mean stress of the basal, middle, and apical segments were 7.7 ± 1.8 kPa, 9.1 ± 3.8 kPa, and 6.4 ± 0.4 kPa, respectively. This is the first patient-specific computational model of LFLG AS based on clinical imaging. Low myocardial stress correlated with low ejection fraction and eccentric LV remodeling. Further studies are needed to understand how alterations in LV biomechanics correlates with clinical outcomes of AS.