TL;DR
This paper introduces a computational fluid dynamics (CFD) model to investigate how peripheral arterial disease (PAD) affects the risk factors for abdominal aortic aneurysms (AAAs). By quantifying uncertainties in key parameters, the study reveals that the amplitude and frequency of blood flow at the inlet are the most critical factors influencing AAA risk, and that ECAP is a more reliable indicator of risk than RRT.
What Problem Does This Paper Address?
Abdominal aortic aneurysms (AAAs) are dangerous dilations of the aorta that can rupture, leading to high mortality. Peripheral arterial disease (PAD), which affects blood flow in the limbs, often coexists with AAAs. While PAD is known to be a risk factor for AAA growth and rupture, the specific haemodynamic mechanisms by which PAD influences AAA risk remain largely unexplored. Accurately simulating blood flow in complex arterial networks, especially with the added complexity of PAD, involves numerous uncertain parameters. This paper aims to address this gap by developing a robust CFD model to quantify the impact of PAD on AAA risk factors and to understand how uncertainties in model parameters propagate to these risk factors.
Methods
The researchers employed a three-dimensional lattice Boltzmann method (LBM) solver, HemeLB, to simulate blood flow within a detailed model of the human abdominal aorta, including an AAA. To represent the effects of PAD, they manipulated the flow rates at the arterial outlets using two-element Windkessel (WK2) models, which are standard for simulating vascular networks.
A core contribution of this work is the application of uncertainty quantification (UQ) and sensitivity analysis to this complex haemodynamic model. Specifically, they used:
- Polynomial Chaos Expansion (PCE): A non-intrusive method to approximate the output quantities (risk factors) as a function of uncertain input parameters. This allows for efficient estimation of statistical moments like mean and variance.
- Sobol' Indices: A variance-based method to determine the relative importance of each input parameter to the overall uncertainty in the output quantities. This helps identify which parameters have the most significant impact on AAA risk factors.
- Coefficient of Variation (CV) and Coefficient of Variation Ratio (CVR): Used to assess the variability of the risk factors and to quantify the amplification of input uncertainty in the output.
They focused on nine critical input parameters, including those related to blood properties, elastic wall properties, inlet flow conditions (Reynolds and Womersley numbers), and outlet boundary conditions. The quantities of interest (QoIs) used to assess AAA risk were Time-Averaged Wall Shear Stress (TAWSS), Oscillatory Shear Index (OSI), Endothelial Cell Activation Potential (ECAP), and Relative Residence Time (RRT).
Datasets
The study utilized a pre-existing arterial model from the Vascular Model Repository, which was derived from a CT angiogram of a 69-year-old male. This model was scaled and modified to represent the abdominal aorta with an AAA. No new datasets were introduced; rather, the focus was on simulating blood flow within this established anatomical model.
Key Insights and Ideas
- PAD's Influence via Flow Rates: The study demonstrates that manipulating outlet flow rates using WK2 models can effectively mimic the haemodynamic impact of PAD. This provides a viable approach for incorporating PAD into CFD simulations.
- ECAP as a Reliable Risk Indicator: While both ECAP and RRT are useful for identifying high-risk regions, ECAP proved to be more reliable and adaptable due to its lower coefficient of variation, suggesting it's a more robust predictor of AAA risk.
- Inlet Flow Dominates AAA Risk: The amplitude and frequency of the inlet blood flow (governed by Reynolds and Womersley numbers) were found to be the most influential parameters on AAA risk factors. This highlights the critical need for accurate characterization of these inlet conditions in patient-specific simulations.
- Uncertainty Amplification: The study found that the uncertainty in the output risk factors was significantly amplified (187%) compared to the uncertainty in the input parameters, underscoring the importance of UQ in this domain.
Key Results
- Flow Rate Control: WK2 models effectively controlled outlet flow rates, though specific outlets required more attention for precise control.
- Risk Factor Distributions: TAWSS was lower in the AAA bulge and higher at the entrance/exit. OSI was higher on the flat surface opposite the bulge. ECAP and RRT showed focal regions of high values on the back side of the AAA.
- CV and CVR: The average CVR of 1.87 indicated a doubling of input uncertainty in the risk factors. ECAP showed the greatest sensitivity to initial values, offering flexibility for matching observed data.
- Sobol' Indices: Reynolds (Re) and Womersley (Wo) numbers were the most significant contributors to the variability of AAA risk factors. Parameters like A, E, m, and YR had marginal effects and could potentially be fixed in future simulations. The time of measurement (Ts) had a minor influence, suggesting the simulations reached a quasi-stationary state.
Limitations and Future Work
The authors acknowledge several limitations:
- Single AAA Shape: The study used only one specific AAA geometry. They note that different AAA shapes can induce different flow patterns, which could influence risk factor distributions.
- Parabolic Inlet Profile: The assumption of a parabolic inlet velocity profile, while common, might not fully capture patient-specific haemodynamics. However, they suggest this effect diminishes downstream.
- WSS-Centric Risk Factors: The analysis focused solely on WSS-derived risk factors. Other indicators like vorticity and helicity, which are associated with AAA rupture, were not included.
- PAD Modelling: The study modeled PAD by altering outlet flow rates. It did not explore scenarios with stenosis, occlusion, or dilation in specific peripheral arteries, which represent more complex PAD conditions.
Future work could involve:
- Investigating the impact of different AAA shapes.
- Exploring patient-specific inlet velocity profiles.
- Including a broader range of risk factors beyond WSS-derived metrics.
- Developing more sophisticated models for PAD to capture a wider spectrum of pathological conditions.
- Extending the analysis to pathological flow rate distributions in peripheral arteries.
Summary updated