Abstract
Vessel size imaging (VSI) to provide a measure of vessel radius in the brain has been demonstrated using an injection of contrast agent. Venous vessel radius imaging of brain microvasculature is also possible by exploiting hypercapnia and hyperoxia. However, these respiratory challenges need external devices like masks, monitors, and gas application. In this study, we employ a breath-hold task to mimic hypercapnia for venous VSI. Breath-hold experiments of brain scans were performed on 14 subjects on a 3-T scanner. Parametric maps of mean venous vessel radius were calculated from the changes in R2* and R2, ΔR2* and ΔR2, respectively, which were measured by simultaneous acquisition of gradient-echo and spin-echo signals using a spin- and gradient-echo (SAGE) echo-planar imaging sequence. In addition, we numerically simulated the expected transverse relaxation in voxels with different vessel radii based on randomly distributed cylinders to eventually obtain vessel size index q (ΔR2*/ΔR2) and associated average vessel radii. With this empirical relation of vessel size index q and vessel radius, the mean measured vessel size index was determined, and venous vessel radii in breath-hold were found to be 7.18 ± 0.49 μm in gray matter and 6.06 ± 0.22 μm in white matter. This study demonstrates the feasibility of venous VSI using a simple breath-hold task. The approach avoids contrast agents and specialized gas delivery, providing a practical alternative for assessing vascular properties. Our results show good agreement with previous hypercapnia- and contrast-based studies, supporting the validity of this noninvasive method.