Abstract
4104
Introduction: The super iterative reconstruction technique is believed to be able to calculate the SUV value of the PET image more accurately, but it is still unclear whether it can help reduce the injected radiation drug dose and increase the PET scanning speed. In this study, we analyzed and evaluates the PET image quality in the super-iterative half-time reconstruction of 68Ga-FAPI PET/MR imaging, so as to explore the functions of super-iterative technology in 68Ga-FAPI PET/MR imaging.
Methods: We retrospectively collected cases of 68Ga-FAPI whole body imaging performed on the integrated PET/MR in our center. The inclusion criteria were as followed: stable breathing during the examination, and PET images were clear with obvious local lesions of high uptake. In the end, 15 cases were included with a total of 27 lesions, and the size of the lesions ranged from 0.3 cm3 to 31.0 cm3. The PET images were reconstructed in three ways using the list data. Full-time conventional reconstruction (interation=2), half-time super-iterative reconstruction (β=350) and half-time conventional reconstruction (interation=2). The 42% threshold segmentation method was used to record the SUVmax, SUVmean and volume of the lesion. The SUVmax, SUVmean and standard deviation of the adjacent background of each lesion were also recorded. Finally, the differences of these parameters were compared statistically between the three reconstruction methods. Furthermore, the parameters L/B ration and SNR for evaluation of images quality were also calculated and compared in the three reconstruction methods. Statistics were performed by SPSS (version 24; IBM). The normal distribution and variance were tested before data comparison. If the conditions were met, the three-group comparison was tested by a matched-pair cluster-randomized trial, otherwise the Friedman nonparametric test was used. p<0.05 is considered to have a significant difference.
Results: Compared to the conventional reconstruction methods with full time, the SUVmax and SUVmean of the lesions were significantly improved in the super iterative reconstruction method (p=0.00), and the background SUVmean was not significantly different (p=0.3). However, the background SUVmax showed a significant increase (p=0.00). For L/B ratios, the value of super-iterative half-time reconstruction was no different both with half-time and full time conventional reconstruction. For the SNR of the image, compared with the full-time reconstruction, the half-time reconstruction image has a significant decrease (p=0.00), but the super iterative half-time reconstruction has a certain increase in the mean SNR compared to the conventional half-time reconstruction. However, There was no significant difference (p=0.29).The detail results were given in Tables in the attached file.
Conclusions: The super iterative PET reconstruction technology can both improve the SUVmax and SUVmean of the lesion, and has no effects for the background noise compared with the conventional PET reconstruction with same condition. Thereby the super iterative PET reconstruction technology has highly advantages for reducing the acquisition time or injection dose to a certain extent.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.