RT Journal Article SR Electronic T1 Comparison of FDDNP and 15O-water PET for cerebral perfusion imaging in patients with traumatic brain injury JF Journal of Nuclear Medicine JO J Nucl Med FD Society of Nuclear Medicine SP 1754 OP 1754 VO 56 IS supplement 3 A1 Koon-Pong Wong A1 Marvin Bergsneider A1 Vladimir Kepe A1 Thomas Glenn A1 Paul Vespa A1 David Hovda A1 Jorge Barrio A1 Sung-Cheng Huang YR 2015 UL http://jnm.snmjournals.org/content/56/supplement_3/1754.abstract AB 1754 Objectives To compare cerebral blood flow (CBF) obtained from 15O-water PET with normalized perfusion obtained from FDDNP PET in patients with traumatic brain injury (TBI).Methods Six TBI patients underwent dynamic 15O-water and FDDNP PET scans sequentially under the same conditions. Summed images of 15O-water (0-10 min p.i.) and FDDNP (0-6 min p.i.) were co-registered to the subject’s T1-weighted MR image, which was warped to the MNI single-subject brain template using the symmetric image normalization method. Regions of interest (ROIs) taken from the automated anatomic labeling atlas were transferred to 15O-water and FDDNP PET images using the transformation from template to PET space. Parametric CBF image was generated from 15O-water kinetics using a 1-compartment model and an arterial input function. Normalized perfusion (R1) image was generated from FDDNP kinetics by simplified reference tissue model using cerebellar gray matter (CGM) as reference tissue. CBF and early-summed FDDNP images were divided by their values in CGM to create normalized CBF (nCBF) and normalized early-summed FDDNP (Rp) images, respectively. Regional values of CBF, nCBF, Rp and R1 were extracted for 19 cortical/subcortical and white matter (WM) ROIs.Results CBF (mean±SD, n=6) in whole-brain-averaged gray matter (WBGM), CGM and WM were 33.1±5.1, 38.4±5.4 and 20.5±4.4 mL/100g/min, respectively. nCBF in WM (0.54±0.11) and WBGM (0.86±0.08) were comparable to those of Rp (0.51±0.09 and 0.92±0.07) and of R1 (0.43±0.09 and 0.88±0.09). Bland-Altman plot showed good agreement of nCBF with Rp (bias=0.05±0.06) and R1 (bias=0.0±0.07) and linear regression slope was ~1. Dunn's multiple comparisons showed comparable nCBF and Rp in 10 of 19 ROIs, with the rest (mainly along anterior, superior and lateral surfaces) having slightly higher Rp due to higher spatial resolution.Conclusions Normalized perfusion estimate provided by FDDNP PET can be used as a surrogate measure of CBF obtained with 15O-water in TBI study.Research Support This work was supported by NIH grants R21-NS053475, P01-NS058489 and P01-AG025831.