Purpose: Cerebral blood flow (CBF), oxygen extraction fraction (OEF) and cerebral metabolic rate of O2 (CMRO2) can be quantified by PET with the administration of H 2 15 O and 15O2. Recently, a shortening in the duration of these measurements was achieved by the sequential administration of dual tracers of 15O2 and H 2 15 O with PET acquisition and integration method (DARG method). A transmission scan is generally required for correcting photon attenuation in advance of PET scan. Although the DARG method can shorten the total study duration to around 30 min, the transmission scan duration has not been optimized and has possibility to shorten its duration. Our aim of this study was to determine the optimal duration for the transmission scan. We introduced 'N-index', which estimates the noise level on an image obtained by subtracting two statistically independent and physiologically equivalent images. The relationship between noise on functional images and duration of the transmission scan was investigated by N-index. Methods: We performed phantom studies to test whether the N-index reflects the pixel noise in a PET image. We also estimated the noise level by the N-index on CBF, OEF and CMRO2 images from DARG method in clinical patients, and investigated an optimal true count of the transmission scan. Results: We found tight correlation between pixel noise and N-index in the phantom study. By investigating relationship between the transmission scan duration and N-index value for the functional images by DARG method, we revealed that the transmission data with true counts of more than 40 Mcounts results in CBF, OEF, and CMRO2 images of reasonable quantitative accuracy and quality. Conclusion: The present study suggests that further shortening of DARG measurement is possible by abridging the transmission scan. The N-index could be used to determine the optimal measurement condition when examining the quality of image.
ASJC Scopus subject areas