Quantitative cerebral perfusion was measured in vivo using dynamic susceptibility contrast magnetic resonance imaging. A dual-echo acquisition was used to eliminate T(1)-enhancement. The arterial input curve was measured in a separate slice in the neck to minimize partial volume effects. Data analysis was performed using a maximum likelihood expectation maximization method to be less sensitive to noise or contrast arrival time differences. From the contrast response curves obtained, the cerebral blood volume (CBV) and flow (CBF) and the timing parameters mean transit time (MTT), time of appearance (TA), and time-to-bolus peak (TBP) were obtained. Adjacent slices were measured to permit discrimination between intra- and inter-subject variance. The group investigated consisted of 41 subjects without cerebral pathology on anatomical MRI. Perfusion parameters for gray (GM) and white matter (WM) were obtained: CBV (GM) = 6.78 +/- 0.99 ml/100 ml, CBV (WM) = 3.78 +/- 0. 96 ml/100 ml, CBF (GM) = 68.7 +/- 21.2 ml/100 ml/min, CBF (WM) = 35. 8 +/- 12.7 ml/100 ml/min, and average GM/WM ratio for CBV (GM/WM) = 1.87 +/- 0.42 and CBF (GM/WM) = 1.99 +/- 0.48. Measured temporal aspects of perfusion were: mean transit time (MTT) (GM) = 6.4 +/- 1. 8 seconds, MTT (WM) = 6.9 +/- 2.3 seconds, time of appearance (TA) (GM) = 1.4 +/- 0.9 seconds, TA (WM) = 2.0 +/- 1.0 seconds, and time-to-bolus peak (TBP) (GM) = 2.4 +/- 1.4 seconds, TBP (WM) = 3.0 +/- 1.5 seconds. The average values were in agreement with those from the literature. Inter- and intra-person variances were estimated using an ANOVA test, and the sources of variance in the parameters, such as image noise, biological variability, and measurement errors of the arterial input curve were found to be of the same order of magnitude. J. Magn. Reson. Imaging 1999;10:109-117.
Copyright 1999 Wiley-Liss, Inc.