Investigation of iterative image reconstruction in low-dose breast CT

Junguo Bian, Kai Yang, John M Boone, Xiao Han, Emil Y. Sidky, Xiaochuan Pan

Research output: Contribution to journalArticle

31 Scopus citations

Abstract

There is interest in developing computed tomography (CT) dedicated to breast-cancer imaging. Because breast tissues are radiation-sensitive, the total radiation exposure in a breast-CT scan is kept low, often comparable to a typical two-view mammography exam, thus resulting in a challenging low-dose-data-reconstruction problem. In recent years, evidence has been found that suggests that iterative reconstruction may yield images of improved quality from low-dose data. In this work, based upon the constrained image total-variation minimization program and its numerical solver, i.e., the adaptive steepest descent-projection onto the convex set (ASD-POCS), we investigate and evaluate iterative image reconstructions from low-dose breast-CT data of patients, with a focus on identifying and determining key reconstruction parameters, devising surrogate utility metrics for characterizing reconstruction quality, and tailoring the program and ASD-POCS to the specific reconstruction task under consideration. The ASD-POCS reconstructions appear to outperform the corresponding clinical FDK reconstructions, in terms of subjective visualization and surrogate utility metrics.

Original languageEnglish (US)
Pages (from-to)2659-2685
Number of pages27
JournalPhysics in Medicine and Biology
Volume59
Issue number11
DOIs
StatePublished - Jun 7 2014

    Fingerprint

Keywords

  • computed tomography
  • image quality assessment
  • optimization-based reconstruction

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging
  • Medicine(all)

Cite this