Trade-off between spatial details and motion artifact in multi-detector CT: A virtual clinical trial with 4D textured human models

Ehsan Abadi, W. Paul Segars, Brian Harrawood, Shobhit Sharma, Thomas Sauer, Anuj Kapadia, Ehsan Samei

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

In Computed Tomography (CT) imaging, high pitch and wide beam collimations accelerate imaging acquisitions and thus reduce motion artifacts. However, increasing pitch and collimation impact acquisition geometry and thus spatial quality of the images. The purpose of this study was to quantify the effects of pitch and beam collimation on image quality using a realistic virtual clinical trial (VCT) construct. The study used extended-cardio torso (XCAT) phantoms enhanced by synthesizing intra-organ heterogeneities within the lungs and bones. Different amounts of cardiac and respiratory motion were simulated, including heart rates of 0, 60, 90, 120 beats per minute, and respiratory rates of 0, 8, 12, 16 breadths per minute. Each case was imaged using a realistic, scanner-specific, and rapid CT simulator setup based on the geometry and physics of a commercial CT scanner (Siemens Definition Flash), at 120 kV, under multiple pitch values, beam collimations, all at the same dose levels. With the knowledge of the ground truth, the quality of the acquired images was quantified by measuring root mean squared error (RMSE) in the lungs. In general, results indicated that RMSE was higher for the phantoms with more respiratory or cardiac motions. For the pitch experiment, images with higher pitch values had less in-plane motion artifacts. RMSE also increased with increasing the pitch. However, the slope of this trend was found to be a function of motion profile, showing the trade-off between motion artifacts and spatial detail loss. For the beam collimation experiment, no major change was observed in in-plane motion artifacts with changes in the beam collimation. RMSE was almost constant with the increase in beam collimation. This study demonstrates the utility of a realistic VCT construct in quantitative evaluation and optimization of CT imaging protocols, when designing such a trial is ethically-prohibitive, costly, and time-consuming.

Original languageEnglish
Title of host publicationMedical Imaging 2019
Subtitle of host publicationPhysics of Medical Imaging
EditorsTaly Gilat Schmidt, Guang-Hong Chen, Hilde Bosmans
PublisherSPIE
ISBN (Electronic)9781510625433
DOIs
StatePublished - 2019
Externally publishedYes
EventMedical Imaging 2019: Physics of Medical Imaging - San Diego, United States
Duration: Feb 17 2019Feb 20 2019

Publication series

NameProgress in Biomedical Optics and Imaging - Proceedings of SPIE
Volume10948
ISSN (Print)1605-7422

Conference

ConferenceMedical Imaging 2019: Physics of Medical Imaging
Country/TerritoryUnited States
CitySan Diego
Period02/17/1902/20/19

Funding

This study was supported by research grants from the National Institutes of Health (R01EB001838 and R01HL131753). Authors would like to thank Karl Stierstorfer, Martin Sedlmair, and Juan Carlos Ramirez from Siemens Healthcare for providing proprietary information which enabled us to model a Siemens CT scanner and for providing us their manufacturer-specific reconstruction software, ReconCT.

Keywords

  • CT simulator
  • Computational phantom
  • Computed Tomography
  • Ray tracer
  • Virtual clinical trial

Fingerprint

Dive into the research topics of 'Trade-off between spatial details and motion artifact in multi-detector CT: A virtual clinical trial with 4D textured human models'. Together they form a unique fingerprint.

Cite this