Comparison of four dual energy image decomposition methods

Abstract

In this study we propose two new methods, namely iso-transmission line and subregion direct approximation methods, for dual energy image decomposition calculations. These two methods are compared with two other conventional techniques, i.e. non-linear equations and direct approximation. The computation accuracy, sensitivity to quantum noise and efficiency are used as indices for comparison. Both the subregion direct approximation method and the iso-transmission line method give better accuracy than the two conventional methods. The iso-transmission line, direct approximation and subregion methods are computationally more efficient than the non-linear equations method. The image calculated by the iso-transmission line and the subregion methods are more sensitive to noise.

An Extended Algebraic Reconstruction Technique (E-ART) for Dual Spectral CT

“…Generally, is selected to be the intersection length of the th X-ray path with the th pixel. With the above notations, we obtain the discrete form of polychromatic projections (6) where…”

“…After discretization, the DSCT reconstruction problem is converted to solving and from (6), which is a nonlinear system.…”

“…However, in dual spectral CT [DSCT, also referred to as dual energy CT (DECT)], the object is scanned under two different X-ray spectra. The DSCT data, which include more information about the scanned object, can be used to reconstruct energy-selective images, images of the electron density and the effective atomic number, or material-selective images [2]- [6]. Therefore the distinguishability of DSCT is remarkably higher than that of standard CT. DSCT has wide potential applications, such as bone mineral density and liver iron concentrations measurements, PET attenuation correction, bone marrow composition analysis, rock core characterization for petrochemical industry, nondestructive evaluation, security inspection, and so forth [7], [8].…”

A new projection-based iterative image reconstruction algorithm for dual-energy computed tomography

“…However, this method would be slow and inefficient. [15,22] In order to solve (8) and (9), we propose a new iterative algorithm that is more robust and accurate than the first-order and the second-order approaches. Downloaded by [Orta Dogu Teknik Universitesi] at 11:01 30 December 2015…”

A GPU-based multi-resolution approach to iterative reconstruction algorithms in x-ray 3D dual spectral computed tomography

“…Basically, these reconstruction methods can be classified into two classes: direct approximation algorithms [11]- [15] and iterative algorithms [16]- [24]. Direct approximation algorithms require neither the knowledge of the x-ray spectrum nor of the energy-dependent attenuation coefficients, and allow a 2 faster material decomposition, but their reconstructed material density images suffer from beam hardening artifacts due to errors introduced in the calibration process of material decomposition.…”