Correlation between SBR and SUV for each correction in the dopamine transporter SPECT

1813 Objectives: Recently, the standardize uptake value (SUV) is validated as the quantitative index in many SPECT examinations. Although the dopamine transporter SPECT is mainly used a specific binding ratio (SBR) as a quantitatively index, the SUV has not fully evaluated. Thus, we have to reveal the correlation of SBR and SUV by each correction method to confirm the utility of SUV. The goal of this study is to reveal correlation between SBR and SUV for each correction in the dopamine transporter SPECT. Method: The 123I solution was filled in two types of striatal phantom such as separated or integrated caudate and putamen. The radioactive concentrations of 123I solution created 36.0, 26.6, 18.0, 13.2 and 4.7 kBq/mL corresponding radioactive ratio of approximately 8:6:4:3:1. We created different phantom of 6 patterns as separating type and 4 patterns as integrating type. SPECT scans used a dual head SPECT/CT system (Discovery NM/CT 670 Q.suit Pro, GE Healthcare Japan, Co., Ltd.) equipped with extended-low-energy general-purpose (ELEGP) collimator. SPECT acquisition was performed using a 128×128 matrix 1.0 zooming, with 120 projections a circular orbit of 15 cm radius, and pixel size of 4.42 mm and total acquisition time was 28-min. The SPECT images were reconstructed using ordered subset expectation maximization (OSEM) incorporating attenuation correction (AC), scatter correction (SC) and resolution recovery (RR). The subset and iteration of OSEM parameter were 10 and 6. Butterworth filter was used as smoothing filter, and a power and cut-off frequency were 16 and 0.45 cycles/cm. We created transverse images of four correction patterns with AC, AC and SC (ACSC), AC and RR (ACRR), and AC,SC and RR (ACSCRR). The SBR and SUVmean were calculated using DaT view (AZE Co., Ltd.) and Q.Metrix (GE Healthcare Co., Ltd.) software programs. True value defined as the radioactive concentration ratio of the well-type scintillation counter, and we compared the true and measured values to reveal the correlation of SBR and SUVmean by each correction Methods: Result: The correlation coefficient of true and measured values for SBR and SUVmean was 0.954 and 0.951 with AC, 0.975 and 0.954 with ACSC, 0.964 and 0.958 with ACRR and 0.971 and 0.960 with ACSCRR, respectively, thereby all correction methods significantly showed excellent positive correlation both SBR and SUVmean (p<0.001). Furthermore, the SBR and SUVmean with ACSC were 12% lower value than those with AC. In addition, the SBR with ACRR was 5% lower value than that with AC, whereas SUVmean with ACRR was 36% higher value than that with AC. Conclusions: The SUVmean could evaluate as quantitative index as well as SBR for dopamine transporter SPECT. Furthermore, the correlation of SBR and SUVmean by each correction method has revealed.