 |
 |
 |
Engineering Optimization of Roller Pump Cardiopulmonary Bypass: High Fidelity In Silico Model of Blood Shear Stress
Yunpeng Tu2, Yi-Ting Yeh2, Alex de Lecea2, Kevin Charette1, Lyubomyr Bohuta1, Vishal Nigam1, Juan Carlos del Álamo2, *Christina L. Greene1
1Seattle Children's Hospital, Seattle, Washington, United States, 2Mechanical Engineering, University of Washington, Seattle, Washington, United States
Objective: Roller pump Cardiopulmonary Bypass (RP-CPB) is the predominant form of pediatric CPB worldwide. The advantages of RP-CPB are that it is less expensive, requires lower priming volume, and has a reliable constant flow rate compared to Centrifugal CPB. The downside is high shear stress on blood causing hemolysis and postoperative inflammation. However, no systematic analysis has been performed on CPB parameters and shear stresses leaving room for optimization. We hypothesize that RP-CPB-associated shear stresses can be mitigated by data-driven optimization of the RP-CPB parameters like blood viscosity (Hct), pump roller occlusion, and rotational speed. This work aims to delineate the effect of the RP-CPB parameters on peak blood shear stress.
Methods: The roller pump is the primary source of the supraphysiologic shear stresses experienced by cells circulating through pediatric CPB circuits. We developed three-dimensional in silico models of this pump and computed the dependence of flow shear stress on CPB parameters to test our hypothesis. Our closed-loop tubing model (Figure 1A) employed finite-element analysis with fluid-structure interaction to reproduce the dynamics of tubing deformation in the pump raceway (Figure 1B) and resolve blood flow inside the tubing (Figure 1C). We performed simulations for varying Hct (20-40%), roller occlusion levels (46-66%), and rotational speeds (38-76 rpm). To quantify shear stresses in each simulation, we computed the von Mises' yield stresses at each point in the tubing and obtained its mean and peak values, as well as its accumulated value along the path followed by circulating cells. Using data from these simulations, we performed a sensitivity analysis investigating how the different parameter combinations affected shear stresses.
Results: Sensitivity analysis from 40 simulations revealed increased peak shear stresses with increasing occlusion, Hct and roller speed. However, peak shear stresses were more sensitive to Hct than the other two parameters. In addition, this dependence was non-linear and steeper than the dependence of oxygen carrying capacity with Hct. Consequently, it was possible to significantly decrease peak shear stresses by decreasing Hct while maintaining oxygen carrying capacity by compensating with roller occlusion and rotational speed (Figure 2). For instance, the red arrow in Figure 2 indicates a 28% decrease in shear stress while increasing oxygen carrying capacity by 10%. Likewise, decreasing Hct and increasing rotational speed, also made it possible to increase oxygen carrying capacity by 50% while keeping shear stress approximately constant (Figure 2 yellow line).
Conclusions: High-fidelity in silico models of RP-CPB offer a data-driven platform to optimize CPB flow, minimize supraphysiologic shear stresses, and palliate post-CPB systemic inflammation. Multi-factorial effects of CPB parameters on the blood flow can be resolved with these models to provide future reference for the operation setup. In our models, Hct reduction by crystalloid hemodilution, concurrent with increasing roller speed or occlusion level, reduces shear stresses in RP-CPB while maintaining the required level of oxygen capacity.
Back to 2025 Abstracts
