MRI-Based Patient-Specific Data for Computational Fluid Dynamics to Predict Hemodynamic Outcome after Surgical Aortic Valve Replacement

Objectives: Choosing the optimal type, size and position for aortic valve prostheses are mandatory for optimal long-term outcome in patients with aortic valve disease. Virtual therapy planning using patient-specific imaging data are a promising new tool for improving individualized patient treatment.

Methods: Ten patients with aortic valve stenosis (8 of 10 with bicuspid aortic valve; 2 of 10 with tricuspid aortic valve), who received surgical aortic valve replacement, underwent pre- and postoperative cardiac magnet resonance imaging (MRI). Preoperative anatomy and 4D flow data were used as inputs for performing virtual therapy planning. A cardiac surgeon together with an engineer performed virtual surgery (i.e., choosing valve size and position) and computational fluid dynamics (CFD) was used for simulating postinterventional outcome after virtual surgery. Virtual surgery simulation results were then compared with real postoperative outcome measured by MRI (maximum pressure gradient, secondary flow degree as a measure of complex blood flow patterns).

Result: Virtual surgery simulation results were comparable to real postoperative outcome. Maximum pressure gradient across the aortic valve was strongly correlated (R ² = 0.80) and blood flow patterns (i.e., secondary flow degree) were not significantly different between virtual simulation and real outcome (mean, 0.94 vs. 0.84, p = 0.182).

Conclusion: Preliminary data suggest valid prediction of hemodynamic outcome by virtual treatment simulation. Interdisciplinary virtual valve replacement combined with CFD has the potential to be used for computer supported decision making for improving individualized patient treatment and long-term patient outcome.