TY - JOUR T1 - Feasibility of carbon-11-dihydrotetrabenazine ligand influx images in a principal component based analysis JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 1411 LP - 1411 VL - 50 IS - supplement 2 AU - Christopher Bogan AU - Yilong Ma AU - Martijn Muller AU - Nicolaas Bohnen Y1 - 2009/05/01 UR - http://jnm.snmjournals.org/content/50/supplement_2/1411.abstract N2 - 1411 Objectives To investigate if ligand influx images obtained from dynamic [C-11] dihydrotetrabenazine (DTBZ) VMAT2 brain PET images will produce a network pattern characteristic of Parkinson’s disease (PD) when processed with a principal component analysis (PCA) based network approach. Methods Thirty subjects (16 patients and 14 normal controls) underwent [C-11] DTBZ VMAT2 brain PET imaging. The rate of [C-11] DTBZ influx was calculated as a surrogate measure of cerebral blood flow by summing the PET images over the first 4 minutes of scanning (K1-weighted [11-C] DTBZ images). A PCA-based network analysis was performed on the K1-weighted [11-C] DTBZ images in an attempt to reproduce a previously validated PD-related covariance pattern (PDRP) obtained in brain FDG glucose metabolic and regional cerebral blood flow PET studies using the SSMPCA toolbox [Spetsieris, 2006, Ma, 2007]. Image pre-processing involved smoothing the K1-weighted [11-C] DTBZ images by a Gaussian kernel at FWMH = 10 mm. Results There was a pattern represented in the second principal component whose expression significantly discriminated PD patients from normal controls (p<0.01). This pattern contained regions which are present in both the PDRP and a previously reported PD related cognitive pattern (PDCP) [Huang, 2007], however, this unique pattern did not match either the PDRP or the PDCP. Conclusions The findings from this exploratory study show the existence of a unique covariance pattern applicable to our study cohort that has the ability to discriminate PD patients from normal controls. These results should be investigated further with a more homogeneous sample. Research Support Study supported by a grant from the Michael J. Fox Foundation. ER -