Some simple rules for contrast, signal-to-noise and resolution in in-line x-ray phase-contrast imaging

CSIRO Materials Science and Engineering, Private Bag 33, Clayton South 3169, Australia.
Optics Express (Impact Factor: 3.49). 04/2008; 16(5):3223-41. DOI: 10.1364/OE.16.003223
Source: PubMed


Simple analytical expressions are derived for the spatial resolution, contrast and signal-to-noise in X-ray projection images of a generic phase edge. The obtained expressions take into account the maximum phase shift generated by the sample and the sharpness of the edge, as well as such parameters of the imaging set-up as the wavelength spectrum and the size of the incoherent source, the source-to-object and object-to-detector distances and the detector resolution. Different asymptotic behavior of the expressions in the cases of large and small Fresnel numbers is demonstrated. The analytical expressions are compared with the results of numerical simulations using Kirchhoff diffraction theory, as well as with experimental X-ray measurements.

15 Reads
  • Source
    • "In the case of linear shift-invariant systems, S 1 is equal to the integral of the system's point-spread function [34]. In the experimental results presented below, the spatial resolution, x, in the images was evaluated according to the approach used, e.g., in [39] [40], i.e. it corresponded to the minimal width of the sharp edges that can be found in the images. "
    [Show abstract] [Hide abstract]
    ABSTRACT: We report the results of a systematic study of phase-contrast x-ray computed tomography in the propagation-based and analyser-based modes using specially designed phantoms and excised breast tissue samples. The study is aimed at the quantitative evaluation and subsequent optimization, with respect to detection of small tumours in breast tissue, of the effects of phase contrast and phase retrieval on key imaging parameters, such as spatial resolution, contrast-to-noise ratio, x-ray dose and a recently proposed 'intrinsic quality' characteristic which combines the image noise with the spatial resolution. We demonstrate that some of the methods evaluated in this work lead to substantial (more than 20-fold) improvement in the contrast-to-noise and intrinsic quality of the reconstructed tomographic images compared with conventional techniques, with the measured characteristics being in good agreement with the corresponding theoretical estimations. This improvement also corresponds to an approximately 400-fold reduction in the x-ray dose, compared with conventional absorption-based tomography, without a loss in the imaging quality. The results of this study confirm and quantify the significant potential benefits achievable in three-dimensional mammography using x-ray phase-contrast imaging and phase-retrieval techniques.
    Journal of Physics D Applied Physics 08/2014; 47(36):365401. DOI:10.1088/0022-3727/47/36/365401 · 2.72 Impact Factor
  • Source
    • "Several experimental and computational developments have been put forward (Guo et al 2011, Luu et al 2011, De Caro et al 2010, 2011, Ren et al 2011, Anastasio et al 2009, 2010, Kashyap et al 2010, Gong et al 2010, Irvine et al 2010, Beltran et al 2010, Olivo and Speller 2009, Meng et al 2009, Chou and Anastasio 2009, 2010). Detailed descriptions of the factors affecting image contrast, signal-to-noise ratio and spatial resolution have been given by Gureyev et al (2008). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Phase-contrast x-ray imaging (PCI) is an innovative method that is sensitive to the refraction of the x-rays in matter. PCI is particularly adapted to visualize weakly absorbing details like those often encountered in biology and medicine. In past years, PCI has become one of the most used imaging methods in laboratory and preclinical studies: its unique characteristics allow high contrast 3D visualization of thick and complex samples even at high spatial resolution. Applications have covered a wide range of pathologies and organs, and are more and more often performed in vivo. Several techniques are now available to exploit and visualize the phase-contrast: propagation- and analyzer-based, crystal and grating interferometry and non-interferometric methods like the coded aperture. In this review, covering the last five years, we will give an overview of the main theoretical and experimental developments and of the important steps performed towards the clinical implementation of PCI.
    Physics in Medicine and Biology 01/2013; 58(1):R1-R35. DOI:10.1088/0031-9155/58/1/R1 · 2.76 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Phase-contrast imaging is one of the most important emerging x-ray imaging techniques. In this work we analyse, from a theoretical point of view, the in-line phase-contrast image formation under general assumptions. The approach is based on wave-optical theory (Fresnel/Kirchoff diffraction integrals) and on the formalism of the mutual coherence function for the evolution of the coherence wavefield properties. Our theoretical model can be applied to phase-contrast imaging realized both by using highly coherent synchrotron radiation and micro-focus x-ray laboratory sources. Thus, the model is suitable for widespread applications, ranging from material science to medical imaging of human body parts. However, it cannot be applied to polychromatic sources, although the validity of the model does not require particularly demanding characteristics of monochromaticity. In addition, for moderate phase gradients, a useful analytical formula of the phase-contrast visibility is derived, based on the a priori knowledge of source size and distance, pixel detector size, defocus distance, material/tissue dielectric susceptibility and characteristic scales of transversal and longitudinal non-uniformities of the material/tissue dielectric susceptibility. Comparisons both with experimental results published by other authors and with simulations based on a Fourier optics approach have been reported, to confirm the validity of the proposed analytical formula
    Physics in Medicine and Biology 11/2008; 53(22):6619-37. DOI:10.1088/0031-9155/53/22/021 · 2.76 Impact Factor
Show more

Preview (2 Sources)

15 Reads
Available from