Premium
A computed tomography implementation of multiple‐image radiography
Author(s) -
Brankov Jovan G.,
Wernick Miles N.,
Yang Yongyi,
Li Jun,
Muehleman Carol,
Zhong Zhong,
Anastasio Mark A.
Publication year - 2006
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.2150788
Subject(s) - computed radiography , tomography , radiography , medical imaging , iterative reconstruction , attenuation , tomographic reconstruction , physics , optics , image quality , computer science , computer vision , radiology , artificial intelligence , medicine , image (mathematics)
Conventional x‐ray computed tomography (CT) produces a single volumetric image that represents the spatially variant linear x‐ray attenuation coefficient of an object. However, in many situations, differences in the x‐ray attenuation properties of soft tissues are very small and difficult to measure in conventional x‐ray imaging. In this work, we investigate an analyzer‐based imaging method, called computed tomography multiple‐image radiography (CT‐MIR), which is a tomographic implementation of the recently proposed multiple‐image radiography method. The CT‐MIR method reconstructs concurrently three physical properties of the object. In addition to x‐ray attenuation, CT‐MIR produces volumetric images that represent the refraction and ultrasmall‐angle scattering properties of the object. These three images can provide a rich description of the object's physical properties that are revealed by the probing x‐ray beam. An imaging model for CT‐MIR that is based on the x‐ray transform of the object properties is established. The CT‐MIR method is demonstrated by use of experimental data acquired at a synchroton radiation imaging beamline, and is compared to the pre‐existing diffraction‐enhanced imaging CT method. We also investigate the merit of an iterative reconstruction method for use with future clinical implementations of CT‐MIR, which we anticipate would be photon limited.