CT QA Revisited in Context of Tube Current Modulation and Iterative Reconstruction
J Winslow*, J Wilson, O Christianson, E Samei, Duke University Medical Center, Durham, NCSU-C-217BCD-3 Sunday 1:30:00 PM - 2:15:00 PM Room: 217BCD
Purpose: Despite their substantial impact on both patient dose and image quality, automatic-tube-current-modulation (ATCM) and iterative reconstruction (IR) are not typically evaluated or monitored in CT quality assurance (CTQA) programs. A new CT phantom and custom software have been incorporated into the CTQA program at our institution. The purpose of this project was to devise a generic testing platform using these newly designed tools, and introduce this platform as part of regular CTQA testing.
Methods: The Mercury Phantom comprises three tapered and four uniform regions of polyethylene (16, 23, 30, and 37cm in diameter), each of which includes four inserts of different materials: air, Polystyrene, Acrylic, and Teflon. Images were acquired using ATCM on three Siemens Flash and three GE 750HD scanners, and iteratively reconstructed (SAFIRE and ASIR). A custom MATLAB software package provided mAs and standard deviation per image, HU, MTF, NPS, and detectability indices for each insert. Results were compared within and across systems, to be used as baseline values for comparison with future system performance testing.
Results: For each scanner model, the absolute difference in mAs for corresponding images of the baseline scans had a mean value less than 2 mAs, with a standard deviation of 1.3 mAs. For different phantom diameters, HU measurements for all four inserts varied little between identical systems or reconstruction algorithms. For each reconstruction algorithm used, detectability indices based upon MTF and NPS measurements were very reproducible, averaging less than 5% difference. Different CT models varied substantially in terms of their mAs profile.
Conclusions: The Mercury Phantom and data analysis technique can be used to benchmark/monitor CT performance, including ATCM and IR, and identify changes over time and across systems. The methodology can be readily implemented across a fleet of CT scanners at a large academic medical center.