Neck Vessel Cross-Sectional Area Measured with MRI: Scan-Rescan Reproducibility for Longitudinal Evaluations.

Abstract

The cross-sectional area (CSA) of common carotid arteries-internal carotid arteries (CCA-ICAs), vertebral arteries (VAs), and internal jugular veins (IJVs) is influenced by aging. However, the neck vessel CSA can be affected by other factors as well, including subject positioning, hydration, and respiration, especially in longitudinal studies. This study aimed to assess scan-rescan reproducibility of CCA-ICAs, VAs, and IJVs CSA measurements in order to evaluate their feasibility for longitudinal CSA assessments, and to apply the segmentation method on a longitudinal pilot dataset. Two set of 2-dimensional neck magnetic resonance angiography (MRA) images were acquired on a 3-T scanner from two separate datasets: 9 healthy individuals (HIs) were scanned 5 days apart (scan-rescan dataset) and 12 HIs were acquired 5 years apart (baseline-follow-up dataset). CCA-ICAs, VAs, and IJVs were segmented along the whole vessel length between C3 and C7 intervertebral spaces. Repeated measure analysis of covariance, adjusted for cervical level and sample, and Wilcoxon signed-rank sum test were used to assess the scan-rescan and baseline-follow-up CSA differences. Intraclass correlation coefficient (ICC) was also computed to evaluate scan-rescan reliability. No significant CSA differences were found for the scan-rescan and baseline-follow-up CSA comparisons, using the whole vessel length or single cervical level measurements. ICC analysis showed good degree of scan-rescan reproducibility (considering whole vessel measures: ICC > .9, P-value < .001 for CCA-ICAs, ICC > .6, P-value < .001 for VAs, and ICC > .7, P-value < .001 for IJVs). Scan-rescan reproducibility of CCA-ICAs, VAs, and IJVs CSA measurements is high, making longitudinal studies feasible.