Icon Legend

This session is not in your schedule.

This session is in your schedule. Click again to remove it.

Back

Cardiovascular Engineering

Evaluating Techniques for Estimating Hepatic Flow for Hemodynamic Assessment of Total Cavopulmonary Connection

(G-261) Evaluating Techniques for Estimating Hepatic Flow for Hemodynamic Assessment of Total Cavopulmonary Connection

Saturday, October 14, 2023

9:30 AM – 10:30 AM PDT

Location: Exhibit Hall - Row G - Poster # 261

Presenting Author(s)

CA

Carter Allen (he/him/his)

Undergraduate Research Assistant
University of Massachusetts Lowell
Fitchburg, Massachusetts, United States

Primary Investigator(s)

ZW

Zhenglun A. Wei

Assistant Professor
University of Massachusetts Lowell, United States

Introduction:: Fontan surgery, a common palliative procedure for patients with congenital single ventricle defects, results in a total cavopulmonary connection (TCPC) which allows for passive return from systemic veins to the lungs. Fontan patient outcomes have been linked to hemodynamic metrics in the TCPC, including hepatic flow distribution (HFD). TCPC hemodynamics are often studied using computational fluid dynamics (CFD) simulations utilizing a model of the TCPC, which excludes the hepatic vein (HV). In these simulations, the HFD calculation typically assumes uniform mixing of HV flow with inferior vena caval (IVC) flow in the Fontan pathway (FP). Recent literature challenged this assumption, demonstrating non-uniform HV flow mixing in the FP, highlighting the importance of including the HVs for accurate HFD calculations. Since patient-specific HV flow measurements are often unavailable, methods for HV flow estimation are needed to include the HVs in simulations. In similar vessels, it has been demonstrated that a diameter (or area)-based “power law” accurately models the distribution of flow at branches, where total flow is distributed to each branching vessel weighted by some power of the vessels’ diameters. The precise exponent used for the power law is based on physical assumptions about the system. The aim of this study is to compare HV flow rates estimated by area-based power laws to patient-specific HV flow rates for use in CFD simulations that numerically assess TCPC hemodynamics.

Materials and Methods::

TCPC power loss (PL) and HFD were compared between simulations using HV flow estimation and simulations using patient-specific HV flow rates. Three different approaches to HV flow estimation were examined, each using an area-based power law with a different exponent. The first approach distributes total HV flow to each HV weighted by their cross-sectional areas under the assumption of constant flow velocity. The second approach distributes flow to each HV by their area to the power of 3/2, assuming constant wall shear stress. The third approach distributes flow to each HV by the square of their areas, following the Hagen-Poiseuille Law. Total HV flow was calculated as the difference between the FP and sub-hepatic IVC flow rates. Three HV-estimated cases were simulated, one for each estimation approach. For comparison, brachiocephalic vein (BV) flow estimation was also performed using the patient-specific flow from the superior vena cava. This led to six more simulation cases: three BV-estimated cases and three combined (BV-&-HV-estimated) cases. The terms ΔPL and ΔHFD were defined to quantify the difference between the hemodynamic metrics calculated by the estimated models and those calculated by the patient-specific model.

Results, Conclusions, and Discussions:: Figures 1 and 2 illustrate the HFD and PL for each patient calculated from each of the HV flow approaches. There is no remarkable difference in PL between any of the HV-estimated models and the patient-specific model, while the difference in HFD was notable ( >5%) for some patients. However, the performance of the power law for HV flow estimation is not uniform in all cases. For example, the area-square-based power law was the best for patients 2 and 3, while the area-based power law outperformed other approaches for patient 4. The HV-estimated models had a negligible (< 1%) mean ΔPL, compared to the relatively higher BV and combined models (4.36% and 5.55%, respectively). The HV models had relatively higher mean ΔHFD values than the BV models. It is possible that discrepancies between the estimated and patient-specific HV flow led to different HV flow mixing patterns in the FP, resulting in a different HFD. Nevertheless, future investigations of the clinical implications and generalization of the findings are warranted.

Acknowledgements (Optional): :

References (Optional): :