A patient-specific cerebral blood flow model

Area of Science:

• Biomedical Engineering
• Computational Fluid Dynamics
• Medical Imaging

Background:

• Complex cerebrovascular diseases necessitate difficult treatment choices.
• Patient-specific mathematical blood flow models can assist in clinical decision-making.
• Simplified models may suffice when high accuracy is not critical.

Purpose of the Study:

• To validate a patient-specific, simplified mathematical blood flow model.
• To assess the model's feasibility for clinical application in cerebrovascular disease treatment planning.

Main Methods:

• MRI and Noninvasive Optimal Vessel Analysis (NOVA) were used to obtain patient-specific vascular data.
• A Hagen-Poiseuille based mathematical model with patient-specific boundary conditions was employed.
• Model validation was performed on 20 healthy subjects, with autoregulatory parameters optimized using data from 10 additional subjects.

Main Results:

• A median percentual flow difference of -3% (IQR -36% to 17%) was observed.
• Regression analysis yielded R² values of 0.71 for absolute flow.
• Bland-Altman analysis indicated a bias of 5% (-70% to 80%) for absolute flow.

Conclusions:

• The validated simplified model demonstrates accuracy potentially suitable for clinical use.
• Model simplifications and validation limitations necessitate further studies in actual patient populations.
• The model shows promise for aiding treatment decisions in complex cerebrovascular cases.