Quantitative CBF measurement without arterial blood sampling by means of high resolution and large field-of-view (FOV) PET and O-15 water

Objectives By means of recently developed high resolution (transverse spatial resolutions of 3.5 mm in FWHM) and large Field-of-View (FOV, 260mm axially) clinical PET scanner, we can identify a radioactivity in internal carotid artery (ICA) during brain scanning. We tested a possibility that ICA time activity curve (TAC) can be used as an input function for quantitative CBF measurement.

Methods Six normal volunteers (3 male and 3 female; mean age, 50.2±2.7 yr) underwent H₂¹⁵O PET of 180 sec data acquisition. Radial artery (RA) TAC was measured by b-detector during continuous arterial blood sampling. Unilateral ICA-TAC was measured in total of 30 frames (2 sec×10 frames + 6 sec×10 frames + 10 sec×10 frames) by reading maximum count within a region of interest. Area under the ICA-TAC from initial rise to peak was correlated with that under the RA-TAC. ICA-TAC of each subject was calibrated using this correlation. Calibrated ICA-TAC based CBF were compared with the RA-TAC based CBF with a correction for tracer delay and dispersion.

Results Area under ICA-TAC (x) was significantly correlated with that under RA-TAC (y) (y=2.80x, r=0.856, p<0.05). Using this calibration factor, ICA-TAC based CBF showed highly significant correlation with RA-TAC based CBF for 24 brain regions (y=1.01x–0.54, r=0.999). The mean difference between RA-TAC based and ICA-TAC based CBF was 0.38%.

Conclusions The present study indicated that arterial blood sampling-based TAC can be substituted by ICA-based TAC by means of high resolution and large FOV PET scanner.