Multi-Objective Optimization of a Ventricular Assist Device Rotor: Balancing Pressure Rise, Efficiency, and Torque

The design and optimization of ventricular assist devices (VADs) are critical for improving their hydraulic performance and clinical outcomes. This study employs computational modeling, feature importance analysis, and response surface optimization to enhance the performance of a VAD rotor. A three-dimensional, steady-state, incompressible flow model was used to simulate blood flow, with blood modeled as a Newtonian fluid. Predictive model for pressure rise (PR), torque (TO), and hydraulic efficiency (EF) was developed and rigorously evaluated, achieving R2 scores of 0.99, 0.99, and 0.95, respectively, indicating high predictive accuracy. Feature importance analysis revealed that rotational speed is the most influential parameter for pressure rise and torque, while inlet blade angle and clearance gap are critical for hydraulic efficiency. Response surface analysis provided insights into the interactions between key design parameters, guiding the optimization process. The optimized design achieved a pressure rise of 36,900 Pa (a 237.2% improvement over the baseline), a torque of 0.005 N·m (a 37.5% decrease), and a hydraulic efficiency of 36.4% (a 29.5% improvement). These results demonstrate the effectiveness of the optimization approach in balancing competing performance objectives. The findings have significant implications for the development of next-generation VADs, offering improved hemodynamic support and energy efficiency, which could enhance patient outcomes and device durability. Future work will focus on experimental validation and further optimization to address limitations and enhance clinical applicability.