Comparative study of the FAIR technique of perfusion quantification with the hydrogen clearance method

Abstract

Arterial spin labeling magnetic resonance methods, including flow-sensitive alternating inversion recovery (FAIR), are becoming increasingly common for the noninvasive quantification of cerebral blood flow (CBF). This report compares the FAIR method with hydrogen clearance. The latter is an established, invasive technique for CBF measurement in animals. Paired readings of CBF were obtained in gerbils to maximize the degree of spatial and temporal correspondence between methods. Flow-sensitive alternating inversion recovery (50 averages, 6.7-minute measurement time) and hydrogen clearance measurements were made concurrently. Cerebral blood flow values measured by both techniques displayed an initial decrease because of the injurious effects of electrode insertion and subsequent recovery. Mixed model regression analysis, structural equations modeling, and a simple concordance correlation coefficient analysis were performed. No evidence of a marked systematic bias in the FAIR measurements was found; mixed model regression analysis yielded relative bias estimates of 0.4 (confidence interval: 3.0, 3.9) mL · 100 g-1 · min-1 and -3.7 (-12.1, 4.7) mL · 100 g-1 · min-1 at 20 and 100 mL · 100 g-1 · min-1, respectively. The principal limitation of the FAIR technique was the magnitude of the random measurement error (imprecision), which had a standard deviation on the order of 10 mL · 100 g-1. min-1.