Test-retest reproducibility of quantitative CBF measurements using FAIR perfusion MRI and acetazolamide challenge

The reproducibility of quantitative cerebral blood flow CBF) measurements using MRI with arterial spin labeling and acetazolamide challenge was assessed in 12 normal subjects, each undergoing the identical experimental procedure on two separate days. CBF was measured on a 1.5T scanner using a flowsensitive alternating inversion recovery (FAIR) pulse sequence, performed both at baseline and 12 min after intravenous administration of acetazolamide. T, was measured in conjunction with the FAIR scan in order to calculate quantitative CBF. The CBF maps were segmented to separate gray matter (GM) from white matter (WM) for region-of-interest (ROI) analyses. Postacetazolamide CBF values (ml/100 g/min, mean +/- SD) of 87.5 +/- 12.5 (GM) and 46.1 +/- 10.8 (WM) represented percent increases of 37.7% +/- 24.4% (GM) and 40.1% +/- 24.4% (WM). Day-to-day differences in baseline CBF were -1.7 +/- 6.9 (GM) and -1.4 +/- 4.7 {WM) or, relative to the mean CBF over both days for each subject, -2.5% +/- 11.7% (GM) and -3.8% +/- 13.6% (WM) Dayto-day differences in absolute post-ACZ CBF increase were -2.5 +/- 6.8 (GM) and 2.7 +/- 9.4 (WM) or, relative to the mean CBF increase over both days for each subject, -4.7% +/- 13.3% (GM) and 9.1% +/- 26.2% (WM). Thus, FAIR-based CBF measurements show satisfactory reproducibility from day to day, but with sufficient variation to warrant caution in interpreting longitudinal data. The hemispheric asymmetry of baseline CBF and post-acetazola mide CBF increases varied within a narrower range and should be sensitive to small changes related to disease or treatment. (C) 2002 Wiley-Liss, Inc.