Simulation model for contrast agent dynamics in brain perfusion scans.

Area of Science:

• Neuroimaging
• Medical Physics
• Computational Biology

Background:

• Quantitative brain perfusion imaging methods exhibit significant heterogeneity.
• Standardization is crucial for reliable and comparable diagnostic outcomes.
• Existing methods lack standardized test data for unbiased comparison.

Purpose of the Study:

• To develop a novel brain perfusion simulation model for generating realistic, unbiased test data.
• To facilitate the comparison and validation of quantitative brain perfusion analysis methods.
• To address the heterogeneity in current perfusion imaging techniques.

Main Methods:

• A simulation model was created using a combination of cerebral arterial and microvascular networks.
• Blood and contrast agent dynamics were modeled using convection-diffusion in tubular networks.
• The model was configured for embolic stroke scenarios, generating arterial input and tissue concentration curves.

Main Results:

• The simulation model produced physiologically plausible vascular dispersion operators (gamma-variate function) and tissue retention functions (boxcar and exponential decay).
• Simulated data, including arterial input and time-concentration curves, were generated for various flow and perfusion statuses.
• The model successfully represented an embolic stroke in a middle cerebral artery territory.

Conclusions:

• The proposed brain perfusion simulation model provides a valuable tool for generating standardized test data.
• This approach enables unbiased assessment of the strengths and weaknesses of different quantitative perfusion analysis methods.
• Implementing physiologically plausible operators enhances the validity of simulation-based validation studies.