Standardization of CT Scanner Performance in Routine Clinical Exams: A Pilot Survey
X Duan*, Y Zhang, S Leng, L Yu, C McCollough, Mayo Clinic, Rochester, MNWE-C-103-7 Wednesday 10:30AM - 12:30PM Room: 103
The goal of this project is to standardize CT scanner performance across scanner models, software versions, and patient sizes. In this pilot study, we measured image quality and radiation dose for several CT systems for 3 patient sizes to determine the amount of variation that exists within our practice for our routine abdomen/pelvis scan protocol parameters.
4 CT scanners used in a single practice were studied. A wire phantom was used to measure the modulation transfer function (MTF), a foil phantom to measure section sensitivity profile (SSP), and semi-anthropomorphic abdomen phantoms of three sizes (15 year old, medium and large adults) to measure noise and assess automatic exposure control (AEC) performance. Scans of the abdomen phantom were performed by a clinical technologist as if the phantom were a patient, with all parameters per our clinical protocols. Wire and foil phantoms were scanned with AEC off using overlapping thin images, but with the same reconstruction kernel.
The 10% values of MTFs were 6.6, 6.4, 6.8 and 6.6 lp/cm and full-widths-at-half-maximum of SSP were 4.9, 5.0, 5.0 and 5.3 mm. The coefficients of variation of recorded CTDIvol were 18%, 20% and 24% for 15 year old, medium adult and large adult phantoms across the four scanners, with noise coefficients of variation of 11%, 6% and 19%.
For routine abdomen scans performed in a single practice using different scanners, there were negligible differences in spatial resolution (MTF or SSP). For a 15 yo and medium adult phantoms, variations of dose and noise were present, but modest. For a large adult, a factor of 1.8 in dose and 1.6 in noise were observed between the lowest and highest dose or noise values, demonstrating an opportunity to adjust our AEC settings to make the dose and noise more consistent across scanners.
Funding Support, Disclosures, and Conflict of Interest: This work was supported in part by NIH grant R01 EB071095 from the National Institute of Biomedical Imaging and Bioengineering.
Add this talk to vcal | ical | Contact Email: