Microvascular Flow Modeling using In Vivo Hemodynamic Measurements in Reconstructed 3D Capillary Networks

OBJECTIVE: We describe a systematic approach to modeling blood flow using reconstructed capillary networks and in vivo hemodynamic measurements. Our goal was to produce flow solutions that represent convective O(2) delivery in vivo. METHODS: Two capillary networks, I and II (84 × 168 × 342 and 70 × 157 × 268 μm(3)), were mapped using custom software. Total network red blood cell supply rate (SR) was calculated from in vivo data and used as a target metric for the flow model. To obtain inlet hematocrits, mass balances were applied recursively from downstream vessels. Pressure differences across the networks were adjusted to achieve target SR. Baseline flow solutions were used as inputs to existing O(2) transport models. To test the impact of flow redistribution, asymmetric flow solutions (Asym) were generated by applying a ± 20% pressure change to network outlets. RESULTS: Asym solutions produced a mean absolute difference in SR per capillary of 27.6 ± 33.3% in network I and 33.2 ± 40.1% in network II vs. baseline. The O(2) transport model calculated mean tissue PO(2) of 28.2 ± 4.8 and 28.1 ± 3.5 mmHg for baseline and 27.6 ± 5.2 and 27.7 ± 3.7 mmHg for Asym. CONCLUSIONS: This outcome illustrates that moderate changes in flow distribution within a capillary network have little impact on tissue PO(2) provided that total SR remains unchanged.