Impact of workflow variability on image-based intracranial aneurysm hemodynamics

Area of Science:

• Biomedical Engineering
• Computational Fluid Dynamics
• Medical Imaging

Background:

• Image-based blood flow quantification aids in identifying high-risk intracranial aneurysms and planning treatments.
• Significant uncertainties exist in the interdisciplinary process, from medical imaging to hemodynamic parameter evaluation.
• Understanding these uncertainties is vital for reliable aneurysm assessment.

Purpose of the Study:

• To systematically analyze uncertainty in hemodynamic simulations of intracranial aneurysms.
• To identify factors influencing simulation accuracy and their interactions.
• To determine pivotal variables for consistent hemodynamic quantification.

Main Methods:

• Varied eight factors affecting hemodynamic simulation accuracy across four patient-specific intracranial aneurysms.
• Performed 1024 transient simulations, evaluating twelve hemodynamic parameters.
• Assessed marginal and interaction effects of influencing factors.

Main Results:

• Global uncertainty analysis showed significant deviations for maximum velocity (Vmax), inflow concentration index (ICI), normalized wall shear stress (WSSnorm), and low shear area (LSA).
• Image reconstruction, inlet, and outlet boundary conditions were identified as major contributors to deviations.
• Rheological modeling and surface smoothing exhibited the lowest median deviations; interaction effects were generally minor.

Conclusions:

• Image reconstruction and boundary conditions are pivotal variables for accurate intracranial aneurysm hemodynamic quantification.
• Minor interaction effects validate the isolated analysis of most influencing factors.
• This study provides essential insights for improving the reliability of image-based hemodynamic analysis in aneurysm research.