In:
Investigative Radiology, Ovid Technologies (Wolters Kluwer Health), Vol. 55, No. 8 ( 2020-8), p. 494-498
Abstract:
The aim of this study was to evaluate the potential of x-ray dark-field radiography for the noninvasive detection of monosodium urate (MSU) crystals as a novel diagnostic tool for gout. Materials and Methods Contrast-to-noise ratios of MSU crystals in conventional radiography and dark-field radiography have been compared in a proof of principle measurement. Monosodium urate crystals have been injected into mouse legs in an ex vivo experimental gout setup. Three radiologists independently evaluated the images for the occurrence of crystal deposits in a blinded study for attenuation images only, dark-field images only, and with both images available for a comprehensive diagnosis. All imaging experiments have been performed at an experimental x-ray dark-field setup with a 3-grating interferometer, a rotating anode tube (50 kVp), and a photon-counting detector (effective pixel size, 166 μm). Results X-ray dark-field radiography provided a strong signal increase for MSU crystals in a physiological buffer solution compared with conventional attenuation radiography with a contrast-to-noise ratio increase from 0.8 to 19.3. Based on conventional attenuation images only, the reader study revealed insufficient diagnostic performance (sensitivity, 11%; specificity, 92%) with poor interrater agreement (Cohen's coefficient κ = 0.031). Based on dark-field images, the sensitivity increased to 100%, specificity remained at 92%, and the interrater agreement increased to κ = 0.904. Combined diagnosis based on both image modalities maximized both sensitivity and specificity to 100% with absolute interrater agreement (κ = 1.000). Conclusions X-ray dark-field radiography enables the detection of MSU crystals in a mouse-based gout model. The simultaneous avaliability of a conventional attenuation image together with the dark-field image provides excellent detection rates of gout deposits with high specificity.
Type of Medium:
Online Resource
ISSN:
1536-0210
,
0020-9996
DOI:
10.1097/RLI.0000000000000671
Language:
English
Publisher:
Ovid Technologies (Wolters Kluwer Health)
Publication Date:
2020
detail.hit.zdb_id:
2041543-6
