X-ray dark-field radiography facilitates the diagnosis of pulmonary fibrosis in a mouse model

Katharina Hellbach, Andre Yaroshenko, Konstantin Willer, Thomas M. Conlon, Margarita B. Braunagel, Sigrid Auweter, Ali Yildirim, Oliver Eickelberg, Franz Pfeiffer, Maximilian F. Reiser, Felix G. Meinel

Research output: Contribution to journalArticlepeer-review

29 Scopus citations

Abstract

The aim of this study was to evaluate whether diagnosing pulmonary fibrosis with projection radiography can be improved by using X-ray dark-field radiograms. Pulmonary X-ray transmission and dark-field images of C57Bl/6N mice, either treated with bleomycin to induce pulmonary fibrosis or PBS to serve as controls, were acquired with a prototype grating-based small-animal scanner. Two blinded readers, both experienced radiologists and familiar with dark-field imaging, had to assess dark-field and transmission images for the absence or presence of fibrosis. Furthermore readers were asked to grade their stage of diagnostic confidence. Histological evaluation of the lungs served as the standard of reference in this study. Both readers showed a notably higher diagnostic confidence when analyzing the dark-field radiographs (p < 0.001). Diagnostic accuracy improved significantly when evaluating the lungs in dark-field images alone (p = 0.02) or in combination with transmission images (p = 0.01) compared to sole analysis of absorption images. Interreader agreement improved from good when assessing only transmission images to excellent when analyzing dark-field images alone or in combination with transmission images. Adding dark-field images to conventional transmission images in a murine model of pulmonary fibrosis leads to an improved diagnosis of this disease on chest radiographs.

Original languageEnglish
Article number340
JournalScientific Reports
Volume7
Issue number1
DOIs
StatePublished - 1 Dec 2017

Fingerprint

Dive into the research topics of 'X-ray dark-field radiography facilitates the diagnosis of pulmonary fibrosis in a mouse model'. Together they form a unique fingerprint.

Cite this