Safire CT iterative reconstruction protocol to reduce patient effective dose on PET/CT examinations

Introduction: To implement the use of CT iterative reconstruction, instead of standard FBP, in order to reduce the patient dose of the 18F-FDG PET/CT examinations, preserving image quality and SUV accuracy. Methods: Catphan 600 and Alderson-Rando anthropomorphic phantom were used to evaluate the SAFIRE iterative CT reconstruction on Siemens Biograph mCT system, progressively reducing the CareDose4D Quality-Ref-mAs. The PET quantitative accuracy was also evaluated by NEMA2012 IQ phantom. The optimized CT reconstruction protocol was applied in clinical setting on additional segmental PET/CT scan, after standard whole body PET/CT, for qualitative evaluation by two nuclear medicine physicians. Results: Measurements on Cathpan phantom demonstrated that Safire reconstruction level 3 allowed to reduce the CT dose of about 40% without worsening the noise and CNR. For mAs-modulated acquisitions on the anthropomorphic phantom, a 30% reduction of dose using Safire at level 3 does not worsen image quality in any anatomical part. Analysing PET/CT acquisitions of NEMA IQ phantom with standard CT FBP reconstruction vs dose-optimized CT iterative reconstruction, the SUV values showed very small deviations (<0.6 %). The clinical evaluation carried out on 25 patients demonstrates the non-inferiority of the image quality obtained with the reduced CT dose and Safire 3 iterative reconstruction protocol compared to the reference CT dose and FBP reconstruction. Conclusions: The use of the Safire iterative reconstruction at level 3 in whole-body PET/CT acquisition allowed an effective dose reduction of 1.8 mSv in mean (from 10.1 mSv to 8.3 mSv, p < 0.001), without any detriment of image quality and SUV accuracy.