Angular super-resolution in X-ray projection radiography using deep neural network: Implementation on rotational angiography

Siow, Tiing Yee; Ma, Chengyu; Toh, Cheng Hong

doi:10.1016/j.bj.2022.01.001

Cited by 4 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This adaptive metric [24] usually combines and focuses on the estimation of contrast, brightness, and structural functions inherent in the image that allows for the adjustment of the respective component's input to the evaluation. As an example, in [25], SSIM was utilized as one of the image quality assessments in the visualization of overlapping vascular networks, where authors built an intermediate data synthesis model that allows them to reduce the required number of projections to lessen the radiation dose absorbed by the head using rotational radiography.…”

Section: Resultsmentioning

confidence: 99%

Interface for Semi-automatic Image Contrast Enhancement in the Propagation-based X-ray Phase Contrast Method

Polishchuk,

Pavlyuk,

Tarasenko

et al. 2023

J. Nano- Electron. Phys.

View full text Add to dashboard Cite

This paper describes a graphical user interface developed for the normalization of radiograms and semiautomatic search for the most contrast images using a given reference, the experimental data for which were obtained using the analytical technique of X-ray phase contrast imaging. The projections of the biological object were obtained by applying free space propagation of coherent radiation when no additional optical elements are placed between the sample and the detector equipped with a YAG:Ce scintillator to exploit the interference effects characteristic of the method due to the high-intensity characteristics of the X-ray light of the SOLEIL synchrotron. The energy of the X-ray photons was 10 keV, and the distance between the sample and the detector was 16.5 cm. We consider the use of standard Matlab tools for algorithmic correction of inhomogeneities of the detector and probing light beam used to build an image of the internal structure of the object and to evaluate the contrast of a group of images, by choosing such common quality criteria as PSNR and SSIM. The program is able to generate such a set of images depending on the parameters entered by the user for a subjective search for contrast images. Numerical estimates well confirm the visual results of the algorithmic search and allow us to reach the most informative image from the generated array, which is created by the user using the interface. The aim of this work is to familiarize with the use of the developed software product and the concept of semi-automatic search among the dataset of digital phase-contrast images.

show abstract

Section: Resultsmentioning

confidence: 99%