Affiliations: [a] Department of Medical Physics and Engineering, Division of Medical Technology and Science, Course of Health Science, Graduate School of Medicine, Osaka University, Osaka, Japan | [b] Department of Radiology, National Center for Child Health and Development, Tokyo, Japan | [c] LISIT Co. Ltd, Tokyo, Japan
Correspondence:
[*]
Corresponding author: Tetsuya Horiuchi, Department of Medical Physics and Engineering, Division of Medical Technology and Science, Course of Health Science, Graduate School of Medicine, Osaka University, 1-7 Yamadaoka, Suita, Osaka 565-0871, Japan. Tel.: +81 6 6879 2571; Fax: +81 6 6879 2571; E-mail: tetsuya.horiuchi@gmail.com.
Abstract: Background:Projection-data based method is established as an accurate simulation of low-dose CT, but cooperation of the CT manufacturer is required. Image-based method is awaited for independent and efficient research. Objective:To develop and verify image-based method for adding noise to DICOM images. Methods:This technique reproduces projection data by fan-beam transform of the DICOM image data and by creating and reconstructing noise projections into a noise image that can be added to the original image. A cylindrical test phantom and anthropomorphic phantom were used to verify the simulation method for both standard and adaptive iterative reconstruction methods. Results:The simulated images compared favorably with actual images. Standard deviation difference of the simulated images were 1.27% ± 0.81% (cylindrical phantom) and ranged from 1.61% ± 0.81% (head), to 5.41% ± 1.57% (hepatic portal region) (anthropomorphic phantom). The standard deviation difference of the cylindrical phantom with the adaptive iterative technique were 3.41% ± 1.76% at 50% ASIR and 3.85% ± 1.82% at 100% ASIR. The simulated noise power spectra of both types of images were comparable to the actual images. Conclusion:We developed and validated a DICOM image-based low-dose simulation that allows researchers to perform dose reduction research independently and efficiently.
