REDUCED TRANSIT-TIME SENSITIVITY IN NONINVASIVE MAGNETIC-RESONANCE-IMAGING OF HUMAN CEREBRAL BLOOD-FLOW

Herein, we present a theoretical framework and experimental methods to more accurately account for transit effects in quantitative human per fusion imaging using endogenous magnetic resonance imaging (MRI) contr ast. The theoretical transit time sensitivities of both continuous and pulsed inversion spin tagging experiments are demonstrated. We propos e introducing a delay following continuous labeling, and demonstrate t heoretically that introduction of a delay dramatically reduces the tra nsit time sensitivity of perfusion imaging. The effects of magnetizati on transfer saturation on this modified continuous labeling experiment are also derived, and the assumption that the perfusion signal reside s entirely within tissue rather than the arterial microvasculature is examined. We present results demonstrating the implementation of the c ontinuous tagging experiment with delay on an echoplanar scanner for m easuring cerebral blood now (CBF) in normal volunteers. By varying the delay, we estimate transit times in the arterial system, values that are necessary for assessing the accuracy of our quantification. The ef fect of uncertainties in the transit time from the tagging plane to th e arterial microvasculature and the transit rime to the tissue itself on the accuracy of perfusion quantification is discussed and found to be small in gray matter but still potentially significant in white mat ter. A novel method far measuring T-1, which is fast, insensitive to c ontamination by cerebrospinal fluid, and compatible with the applicati on of magnetization transfer saturation, is also presented. The method s are combined to produce quantitative maps of resting and hypercarbic CBF.