RT Journal Article SR Electronic T1 Effects of different MR pulse sequences for standardized uptake value (SUV) in dynamic reconstruction of PET in integrated simultaneous PET/MR JF Journal of Nuclear Medicine JO J Nucl Med FD Society of Nuclear Medicine SP 2027 OP 2027 VO 60 IS supplement 1 A1 Zhongwei Chen A1 Bo Wang A1 Changping Liu A1 Wei Zhai A1 Hongzan Sun A1 Jun Xin YR 2019 UL http://jnm.snmjournals.org/content/60/supplement_1/2027.abstract AB 2027Objectives: To analyze the influence of different MR pulse sequences on PET uptake, the standardized uptake value (SUV) values of several dynamic reconstructed PET images acquired simultaneously with different MR sequences in integrated ToF PET/MR were measured and estimated. Methods: Fifteen cases of uterine malignant tumors were retrospectively collected for integrated simultaneous PET/MR (GE Healthcare, WI, USA) pelvic scanning. According to the scan time of different MR sequences, PET images were reconstructed dynamically in exactly the same time as the MR sequences (Fig.1). The maximum and mean SUV (SUVmax and SUVmean) of lesions and muscle background of the reconstructed segmented PET images corresponding to the MR scanning sequences were measured. In order to avoid the influence of time factor for PET reconstruction, segmented PET images related to each MR sequence were reconstructed according to the shortest scanning time (78s) and compared(Fig.2). Single factor ANOVA analysis and post-hoc tests were applied for the difference significance detecting. Furthermore, all PET image were subjectively evaluated by two experienced nuclear medicine diagnosticians according to the five-point grading levels. Results: A representative patient with four MR sequences, their relative simultaneously acquired segmented PET, and fused PET/MR images were shown in Fig.3. After dynamic reconstruction, SUVmax and SUVmean of the lesions and muscle background of the segmented PET images were recorded and compared (Fig.4). ANOVA analysis showed that there was no significant difference between each group of segmented PET images related to each MR sequence in normal and lesion tissues (SUVmax for lesion P=0.076, SUVmean for lesion P=0.174, SUVmax for background P=0.184, and SUVmean for background P=0.419). There were significant differences (all P<0.05) in SUVmax and SUVmean between dynamic reconstructed segmented PET and the whole PET. While in the background muscle, SUVmax, SUVmean of segmented PET had no significant difference (P > 0.05) with the whole PET. In addition, there was no significant difference in SUVmax and SUVmean between the two PET images based on short-time reconstruction with the same MR dynamic reconstruction (All P > 0.05), and therefore time factor for SUV was not considered in this study. According to the international standard principle of five-point grading levels, all segmented and whole PET images of the two physicians for the diagnosis purpose were all above 4 points. and there was no significant difference (P>0.05) between them, which revealed that all PET images involved and reconstructed in this study can be applied for clinical assessment. Conclusions: Dynamic segmented PET qualification related to different MR sequences in the lesion can’t be replaced by the whole PET, which revealed that it is more accurate in dynamic reconstruction. While the SUV in the background location can be replaced by the SUV of the whole PET. However, the segmented PET and the whole PET can be replaced by each other in clinical application, not qualification of lesion. Figure Legends: Fig.1 Integrated PET/MR acquired PET and MR data simultaneously, and segmented PET was dynamic-reconstructed relative to each MR sequence. Fig.2 To avoid time factor, all segmented PET corresponding to the MR sequences were reconstructed as the shortest segmented PET (78s), MR attenuation correction sequence was not considered. Fig.3 A representative patient with four MR sequences, their relative simultaneously acquired segmented PET, and fused PET/MR images were shown. Fig.4 The lesion and background SUVmax of segmented PET images corresponding to MR scanning sequence were shown in Figure 4A and B. While SUVmean of lesion and background were displayed in 4C and 4D.