Validation of dual-source single-tube reconstruction as a method to obtain half-dose images to evaluate radiation dose and noise reduction: phantom and human assessment using CT colonography and sinogram-affirmed iterative reconstruction (SAFIRE)

J Comput Assist Tomogr. 2012 Sep-Oct;36(5):560-9. doi: 10.1097/RCT.0b013e318263cc1b.

Abstract

Objective: To evaluate a method for obtaining half-dose CT images for observer studies evaluating lower-dose CT.

Methods: Phantoms of varying sizes were scanned at multiple tube potentials using dose-matched dual-source (DS) and single-source (SS) protocols. Images from single-tube reconstruction of DS data were compared with SS images acquired at half-original CTDIvol. Thirty patients underwent supine SS and dose-matched prone DS CT colonography (CTC). Half-dose prone images were reconstructed with sinogram-affirmed iterative reconstruction (SAFIRE). Two radiologists scored image quality on 2-dimensional (2D) and 3D images.

Results: Image noise was similar between half-dose SS images and DS images reconstructed from one tube only with tube potential of 120 kV or more for phantoms 40 cm or smaller (P < 0.05). For both readers, the patients' CTC image quality scores were more than 84% concordant between SS or DS CTC images, and half-dose-prone CTC images with SAFIRE had 84% or more concordance with routine-dose CTC except for 3D image noise.

Conclusions: In appropriately sized patients, DS acquisition with single-tube reconstruction can create half-dose images, permitting comparison to full-dose images. For CTC, there is comparable image quality for colonic evaluation between full-dose and half-dose images reconstructed with SAFIRE.

Publication types

  • Research Support, Non-U.S. Gov't
  • Validation Study

MeSH terms

  • Colonography, Computed Tomographic / methods*
  • Female
  • Humans
  • Imaging, Three-Dimensional
  • Male
  • Phantoms, Imaging
  • Radiation Dosage*
  • Radiographic Image Interpretation, Computer-Assisted / methods*
  • Signal-To-Noise Ratio
  • Statistics, Nonparametric