[Computer-assisted decision-making in cardiac surgery: from 3D preoperative planning to computational fluid dynamics in the design of surgical procedures].

Introduction: Although surgical specialties utilize static models for preoperative planning, the evolution of dynamic planning methods and computer simulations created the opportunity for the introduction of dynamic parameters in cardiac surgery. Our aim was to apply 3D models in cardiac surgical practice, predicting fluid dynamic results, ventricular shape, volume and function before the operation.

Methods: Using a script developed by us, the raw DICOM files were imported, the dilated left ventricle was modeled and fluid dynamic parameters were simulated, such as flow kinematic and profile analysis, turbulence calculation and myocardial response to shear stress. Then step-by-step simulation of the surgical ventricle restoration procedure was accomplished and the calculated variables were imbedded in silico model. The length of resection lines was modified based on the previous computer simulation and applied during the operation, if feasible.

Results: The sphericity and conicity indexes were improved significantly in postoperative period (0.42 vs. 0.67 and 0.36 vs. 0.72, p < 0.05). The occurred shear stress at endocardium decreased 83% due to the normalization of flow kinematic pattern of the ventricle in postoperative period (132.21 ± 29.5 dyne/cm vs. 22.92 ± 10.3 p < 0.05 dyne/‌cm). The postoperative turbulent flow pattern significantly decreased, according to our computational method (2712 vs. 1823, p < 0.0001).

Conclusion: With our method, the standardization of the surgical ventricle reconstruction was achievable and the surgical steps were predictable. Therefore, a new decision making support system was established in cardiac surgery for high risk patients. A personalized surgical technique was offered to our patients, improving their life expectancy and quality of life.