PET | Reducing Radiation Exposure using Iterative Statistical CT Reconstruct…

1Yong Ho Do, 1Ho Jun Song, 1Hyung Jin Lee, 1Hong Jae Lee, 1Jin Eui Kim, 1,2Jae Sung Lee, 1,2Dong Soo Lee

1Dept. of Nuclear Medicine, Seoul National Univ. Hospital

2Dept. of Nuclear Medicine, Seoul National Univ.

Purpose: The purpose of this study was to evaluate the effectiveness of iterative CT reconstruction on the radiation dose reduction in PET/CT studies, and to optimize the parameter (kernel size) of the CT reconstruction.

Method: The PET/CT scanner used in this study was Siemens Biograph mCT64, and iterative CT reconstruction software was SAFIRE. An AAPM CT performance phantom was scanned seven times using standard CT parameters (120 kVp, 100 mAs) and reduced tube currents in the step of 15 mAs. After image reconstruction using FBP and SAFIRE, noise and spatial resolution were evaluated. The same methods were applied to an anthropomorphic chest phantom to compare image quality. The same patient was scanned with FBP 40 mAs and SAFIRE 25 mAs for evaluating radiation exposure and image quality. A NEMA IEC

body phantom was used for SUV evaluation. CT dose condition (120 kVp, 40 mAs) was the same for each scan and PET scan durations were 1, 2, 3 and 4 min. After scanning, a different kernel of SAFIRE was applied to ACCT. PET data were reconstructed by each ACCT that adopted SAFIRE.

Results: In the AAPM phantom test for noise evaluation, the SD (standard deviation) in FBP

100 mAs images was 10.5 and SAFIRE 40 mAs was 10.4. The SD of FBP 40 mAs was 15.8, of SAFIRE 25 mAs 15.9, of FBP 25 mAs 22.9, and SAFIRE 10 mAs yielded an SD of 22.9. In the spatial resolution and chest phantom tests there was no significant difference in image quality. In the patient study, SDs for FBP 40 mAs and SAFIRE 25 mAs were 22.4 and 20.0 in muscle, 39.8 and 38.2 in lung and 13.0 and 7.5 in BKG area. In spite of reducing 37.5% of mAs, there was no significant difference in image quality: CTDIvol (mGy) and DLP

(mGy× cm) were decreased 37.5% and 35.5%. Except for I70 kernel, there was no notable difference in the SUV comparison.

Conclusion: It was possible to reduce radiation dose and acquire better image quality than FBPs without SUV change by applying SAFIRE with the appropriate kernel size in PET/CT studies.

Key Words : FBP, SAFIRE, SUV, kernel