Rapid measurement of regional cerebral blood flow in the baboon using 15O-labelled water and dynamic positron emission tomography.

The sensitivity and reproducibility of rapid measurements of regional cerebral flow (rCBF) using a bolus injection of H2 15O and dynamic positron emission tomography (PET) were investigated in anaesthetised baboons. The cerebrovascular reactivity to changes in arterial pCO2 was used as an experimental support. PET data were acquired over 4 min following a single bolus intravenous injection of H2 15O, while arterial blood was withdrawn for continuous activity counting. Images were reconstructed with a dynamic sequence of 45 x 2s + 15 x 10s, including a correction for decay. Regional values of CBF were derived from non-linear least-squares fits of the time activity curves using a four-parameter two-compartment model. The results obtained with a four-parameter fitting method were compared with those obtained with two other rapid estimation methods, first fitting two parameters only, CBF and partition coefficient (p), and secondly autoradiography (with p fixed at 0.95 ml brain ml blood-1). Twelve regions of interest were analysed. The values for the basal CBF obtained from 13 measurements in two baboons were close to published values obtained with other techniques. Reproducibility checks showed a mean variation of 9.7 per cent. The CBF measurements performed in hypercapnic conditions gave results similar to published data in other animal species, showing a 4.5 +/- 0.9 per cent increase in CBF per mm Hg paCO2. The results obtained with the three estimation techniques were closely correlated. The dynamic bolus H2 15O method appeared to be suitable for high blood flow measurements.