Cell accumulation, trapping, and oscillation in a patient-specific microcirculation network under flow*

3D simulations on blood flow in a complex patient-specific retina vascular network were performed considering a mixture of red blood cells (RBCs), white blood cells (WBCs), and obstructed vessels. Without cells, it showed that a vessel blockage in the network might change the flow or even reverse the flow direction on distant vessels. The flow rate in some vessels could increase up to 1200% due to an obstruction. However, with cells, it showed a fluctuating flow pattern, and the cells showed complicated transport behavior at bifurcations. Cell accumulation might occur in some bifurcations such a T shaped junction. The addition of large size of WBCs reduced the local flow rate when they were squeezed through a capillary vessel, resulting a 32% flow rate reduction. The simulation of flow under stenosis with cells showed that cells could oscillate and become trapped in a vessel due to the fluctuating flow. Finally, a reduced order model (ROM) with multiple non-Newtonian viscosity models was used to simulate the blood flow in the network. Among them the Fahræus-Lindqvist model was found to be the most accurate one in terms of predicting the average hematocrit and flow rate in the network, which can be used to build a multiscale model for blood flow.