Noninvasive quantification of cerebrovascular pressure changes using 4D Flow MRI

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Fluid Dynamics

Background:

• Hemodynamic alterations are key indicators of cerebrovascular disease.
• Assessing relative pressure in narrow, tortuous cerebrovasculature using imaging is challenging.

Purpose of the Study:

• To systematically evaluate image-based cerebrovascular relative pressure mapping methods.
• To investigate the accuracy of reduced Bernoulli (RB), extended unsteady Bernoulli (UB), and full-field virtual work-energy relative pressure (WERP) methods.

Main Methods:

• Patient-specific in silico models generated synthetic cerebrovascular 4D Flow MRI data.
• RB, UB, and WERP performance were quantified based on spatiotemporal sampling and image noise.
• Cerebrovascular relative pressures were derived from 4D Flow MRI of healthy volunteers at 1.1 mm and 0.8 mm spatial resolutions.

Main Results:

• In silico analysis showed accurate relative pressure estimation is coupled to spatial sampling; high errors occurred at dx = 1.0 mm for all methods.
• At dx = 0.5 mm, WERP achieved a mean error of 0.02 ± 0.25 mm Hg, significantly lower than RB and UB.
• In vivo data indicated WERP had higher correlative dependence between resolutions (R² = 0.81) than RB (R² = 0.03) or UB (R² = 0.07).

Conclusions:

• Image-based full-field methods like WERP can map cerebrovascular relative pressure.
• The accuracy of WERP is directly dependent on the utilized spatial resolution.
• High spatial resolution is crucial for accurate relative pressure mapping in cerebrovascular imaging.