A tensor-based measure for estimating blood damage

Abstract

Implantable ventricular assist devices give hope of a permanent clinical solution to heart failure. These devices, both pulsatile- and continuous-flow, are presently used as medium-term bridge to heart transplant or recovery. While long-term use of continuous-flow axial and centrifugal pumps is being explored, the excessive level of blood damage in these devices has emerged as a design challenge. Blood damage depends both on shear stress and exposure time, and device designers have relied traditionally on global space- and time-averaged estimates from experimental studies to make design decisions. Measuring distributions of shear stress levels and the blood cell's exposure to these conditions in complex rotary pump flow is difficult. On the other hand, computational fluid dynamics (CFD) is now being used as a tool for designing viable devices, offering more detailed information about the flow field. A tensor-based blood damage model for CFD analysis is proposed here. The model estimates the time- and space-dependent strain experienced by individual blood cells and correlates it to blood damage data from steady shear flow experiments. The blood cells are modeled as deforming droplets and their deformation is tracked along the pathlines of a computed flow. The model predicts that blood cells in a rapidly fluctuating shear flow can sustain high shear stress levels for very short exposure time without deforming considerably. In the context of mechanical modeling of the implantable Gyro blood pump being developed at Baylor College of Medicine, this suggests that blood cells traversing regions of highly fluctuating shear stress rapidly may not hemolyze significantly.