Quantitative cerebral blood flow with bolus tracking perfusion MRI: measurements in porcine model and comparison with H(2)(15)O PET.

ABSTRACT

PURPOSE: We recently presented a method for the quantitative measurement of the arterial input function which allows for determination of absolute cerebral blood flow (CBF) values without adjustable parameters. The aim of the present work is to estimate absolute CBF values by using this new technique and to compare it with the gold standard for cerebral perfusion, H(2)(15)O positron emission tomography. METHODS: Pigs (13) were comparatively investigated by each method performing multiple measurement runs. The reproducibility of both methods was assessed by a voxel-wise correlation of repeated measurements. An intersubject evaluation was performed on median whole-brain CBF estimates. RESULTS: The mean CBF (MRI) was 20±4mL/100g/min for gray matter, the mean CBF (positron emission tomography) was 24±6mL/100g/min for gray and white matter. The reproducibility for MRI correlated with r = 0.85 and P<0.0001, for positron emission tomography with r = 0.76 and P<0.0001. The correlation for the median whole-brain CBF in MRI and positron emission tomography was r = 0.60 and P = 0.04. CONCLUSIONS: The proposed method allows for determination of quantitative CBF without normalization factors. The relatively low estimates of absolute CBF most likely results from the higher age of the pigs as compared to other studies. The intermediate correlation between both methods is caused by physiological intraindividual fluctuations of the CBF and by a limited reproducibility of both methods.