Room: AAPM ePoster Library
Purpose: To validate a normal resolution (NR) simulation algorithm (NRsim) using high resolution (HR) or super high resolution (SHR) physical CT acquisitions on a commercial high resolution CT scanner and compare image quality between NRsim generated images and physical NR acquired CT images. NRsim is intended to allow direct comparison between normal resolution CT and HR/SHR reconstructions in clinical trials, without repeating exams.
Methods: The Aquilion Precision high resolution CT scanner (Canon Medical Systems, Otawara, Japan) has three resolution modes resulting from detector binning in the channel and row directions. For NR, the detector elements are 0.5 mm × 0.5 mm, 0.25 × 0.50 mm for HR, and 0.25 ×0.25 mm for SHR. The NRsim algorithm simulates NR acquisitions from HR or SHR acquisitions (NR_HR and NR_SHR, respectively) by averaging the post-log raw projection data and then reconstructing the down-sampled data using the same process as physical NR acquisitions. Measurements of MTF, NPS, HU accuracy, and low contrast detectability (LCD) were made in phantoms for standard protocols (head, lung, and body) and compared for NR, NR_HR, and NR_SHR images across a range of CTDIvol levels and both FBP and iterative reconstructions.
Results: MTF_10% differences were < 6% across a range of contrast levels and kernels. Differences in noise texture (peak frequency of radial profiles through slices of 3D NPS) were negligible, and the mean absolute difference in noise magnitude (integral of the 3D NPS) was 3.4% and 5.4% for NR_HR and NR_SHR, respectively, across all comparisons. CT numbers were within the expected HU range determined by quantitative modules in a CT phantom. LCD estimations were closely related with a relative RMSE <7.7% for the body protocol.
Conclusion: NRsim generates images with spatial resolution, noise, HU accuracy, and low contrast detectability largely equivalent to images generated using physical NR acquisitions.
Funding Support, Disclosures, and Conflict of Interest: The authors of this abstract declare relationships with Canon Medical Systems Corporation. UC Davis has received funding from Canon Medical Systems Corporation.