RT Journal Article
SR Electronic
T1 The image quality evaluation with various maximum unit difference of the first total-body uEXPLORER scanner
JF Journal of Nuclear Medicine
JO J Nucl Med
FD Society of Nuclear Medicine
SP 389
OP 389
VO 60
IS supplement 1
A1 Zilin Deng
A1 Yu Ding
A1 Yang Lv
A1 Yun Dong
YR 2019
UL http://jnm.snmjournals.org/content/60/supplement_1/389.abstract
AB 389Objectives: The world first total-body uEXPLORER scanner, which is produced by Shanghai United Imaging Healthcare, can be able to provide the clinical imaging in a single breath hold, at 2.9-mm spatial resolution, with high sensitivity about 190 cps/kBq and low dose injection (one tenth of the normal dose). In uEXPLORER, the PET mainly consists 8 units in axial direction. Each unit can do local coincidence within its own unit and also cross coincidence with adjacent units that unit difference ≤4. The true, scatter and random which are dependent with unit difference may cause various effect on image. The aim of this study is to present the image quality with various maximum unit difference (MUD) of the scanner. Methods: The MUD of 1, 2, 3, 4 was chosen for the following investigation. Other common reconstruction parameters include subsets = 20, reconstruction matrix = 256×256, reconstructed FoV = 600 mm. A uniform phantom with 150 mm in diameter and 2000 mm in length was adopted and filled with 18F to study the noise along the axial FoV of the uEXPLORER with MUD first. Then, we characterized the contrast recovery coefficient (CRC) and noise with iterations and each MUD, mainly according to the National Electrical Manufacturers Association (NEMA) NU 2-2012 standard. Furthermore, a 65 kg, 61 years old male volunteer was studied in the clinical evaluation. The volunteer was injected with 7.85 mCi FDG and then scanned under the standard imaging protocol for about 20 minutes following about 60-minute uptake period. A series of PET clinical images were obtained with MUD and the noise was quantified. Results: Figure 1(a) shows the images of the uniform phantom and a region of interest (ROI) was selected on each transverse image slice for noise analysis. Fig. 1(b) presents the noise along the axial FoV with each MUD based on the ROI in fig .1(a). It suggests that the difference of the noise between MUD 3 and MUD 4 is small. The central transverse image slices of the NEMA IEC body phantom (4:1) with various MUD were given in fig. 2 and the corresponding CRC and noise as a functions of iteration were depicted in fig. 3. Fig. 4 reports the results of the clinical images for different MUD and the noise evaluation with a background ROI was shown in table 1. Conclusions: We have demonstrated that the noise decreased with the increase of MUD both in the phantom and clinical study. The convergence may be different due to the random with various MUD. This indicates the reconstructions parameters and MUD need to be optimized with random fraction or NECR for further clinical imaging.