Computer-aided detection for ultra-low-dose CT colonography

  • Authors:

  • Janne J. Näppi;Masanori Imuta;Yasuyuki Yamashita;Hiroyuki Yoshida

  • Affiliations:

  • 3D Imaging Research, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts;Department of Diagnostic Radiology, Kumamoto University, Kumamoto, Japan;Department of Diagnostic Radiology, Kumamoto University, Kumamoto, Japan;3D Imaging Research, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts

  • Venue:
  • MICCAI'12 Proceedings of the 4th international conference on Abdominal Imaging: computational and clinical applications
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

To screen large populations for colorectal cancer, it may be necessary to reduce the radiation dose of computed tomographic colonography (CTC) examinations. We compared the accuracy of computer-aided detection (CAD) in standard-dose (SD) CTC with that in ultra-low-dose (ULD) CTC. We also assessed the effect of linear and nonlinear denoising methods on CAD performance in ULD CTC. The CAD system was trained to detect polyps with 43 SD CTC studies. It was tested with 24 clinical studies, where the supine series were acquired with SD CTC and the prone series were acquired with ULD CTC. The polyp detection accuracy of CAD was significantly lower in ULD CTC than in SD CTC. Linear denoising of ULD CTC images improved the detection accuracy for large polyps, but it reduced sensitivity for small polyps. However, with nonlinear denoising, the detection accuracy of CAD in ULD CTC was not significantly different from that in SD CTC.