Piecewise structural diffusion defined on shape index for noise reduction in dual-energy CT images

  • Authors:

  • Wenli Cai;June-Goo Lee;Da Zhang;Christina Piel;Hiroyuki Yoshida

  • Affiliations:

  • 3D Imaging Research, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts;3D Imaging Research, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA,Department of Radiology, University of Pittsburgh, USA,FARP/Imag ...;3D Imaging Research, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts;3D Imaging Research, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA,Institut für Medizinische Physik und Strahlenschutz, Univer ...;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

The increasing radiation dose in dual-energy CT (DE-CT) scanning due to the double exposures at 80 kVp and 140 kVp is a major concern in the application of DE-CT. This paper presents a novel image-space denoising method, called piecewise structural diffusion (PSD), for the reduction of noise in low-dose DE-CT images. Three principle structures (plate, ridge, and cap) and their corresponding diffusion tensors are formulated based on the eigenvalues of a Hessian matrix. The local diffusion tensor that is piecewise-defined on the domain of shape index is composed by a linear combination of two diffusion tensors of the associated principle structures. A single diffusion tensor calculated from the fused DE-CT image is applied to both high- and low-energy images. In the DE-CT colon phantom study, we demonstrated that DE-CT images filtered by PSD yielded the similar image quality with half of radiation doses.