Normal cerebral perfusion measurements using arterial spin labeling: reproducibility, stability, and age and gender effects.

Before meaningful conclusions can be drawn from clinical measures of cerebral blood perfusion, the precision of the measurement must be determined and set in the context of inter- and intrasubject sources of variability. This work establishes the reproducibility of perfusion measurements using the noninvasive MRI technique of continuous arterial spin labeling (CASL). Perfusion was measured in 34 healthy normal subjects. Intersubject variability was assessed, and age and gender contributions were estimated. Intersubject variation was found to be large, with up to 100% perfusion difference for subjects of the same age and gender. Repeated measurements in one subject showed that perfusion remains remarkably stable in the short term when compared with intersubject variation and the large capacity for perfusion change in the brain. A significant decrease in the ratio of gray-matter to white-matter perfusion was found with increasing age (0.79% per year (P < 0.0005)). This appears to be due mainly to a reduction in gray-matter perfusion, which was found to decrease by 0.45% per year (P = 0.04). Regional analysis suggested that the gray-matter age-related changes were predominantly localized in the frontal cortex. Whole-brain perfusion was 13% higher (P = 0.02) in females compared to males.