The Effect of Extrinsic Material and Radiation Dose on Model-Based Iterative Reconstruction Chest Computed Tomography Reconstruction Time: A Phantom-Based Study

Geoff Karjala; David Leswick; Troy Anderson; Lori Toews; Leslie Chatterson; Chel Hee Lee; Hyun J Lim

doi:10.1016/j.jmir.2017.04.001

The Effect of Extrinsic Material and Radiation Dose on Model-Based Iterative Reconstruction Chest Computed Tomography Reconstruction Time: A Phantom-Based Study

J Med Imaging Radiat Sci. 2017 Sep;48(3):254-258. doi: 10.1016/j.jmir.2017.04.001. Epub 2017 May 25.

Authors

Geoff Karjala¹, David Leswick¹, Troy Anderson¹, Lori Toews¹, Leslie Chatterson², Chel Hee Lee³, Hyun J Lim³

Affiliations

¹ Department of Medical Imaging, University of Saskatchewan, Saskatoon, Saskatchewan, Canada.
² Department of Medical Imaging, University of Saskatchewan, Saskatoon, Saskatchewan, Canada. Electronic address: lch088@mail.usask.ca.
³ Community Health and Epidemiology, College of Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada.

PMID: 31047407
DOI: 10.1016/j.jmir.2017.04.001

Abstract

Objective: Compare effect of extrinsic materials and radiation dose levels on image processing times for model-based iterative reconstruction (MBIR) in computed tomography.

Methods: Chest computed tomography scans were performed on a phantom with three different levels of clothing and medical equipment at three tube current settings to reflect differing radiation doses. Reconstruction time for MBIR was recorded, and objective image quality was assessed via noise within the phantom mediastinum. Reconstruction time and noise were compared between scans, with noise also compared between MBIR and matching filtered back projection (FBP) images.

Results: Reconstruction times (minutes:seconds) ranged from 37:31 to 42:24. Times were generally faster with less extrinsic material and prolonged among high-dose scans when materials were present. On both the MBIR and FBP images, noise levels were improved with higher radiation doses, although for MBIR only minimally, and the relative effect of extrinsic materials at a given radiation dose was also minimal. In addition, noise was better with MBIR than FBP reconstruction for all conditions.

Conclusions: Typical MBIR reconstruction times are faster with less extrinsic materials in the scan field of views, and removing extraneous blankets or medical devices could positively affect workflow over the course of the day. In addition, MBIR reconstruction times are also shorter when using lower dose protocols in situations requiring extensive materials.