Deformable image registration of dark-field chest radiographs for functional lung assessment.

Background: Dark-field radiography of the human chest has been demonstrated to have promising potential for the analysis of the lung microstructure and the diagnosis of respiratory diseases. However, most previous studies of dark-field chest radiographs evaluated the lung signal only in the inspiratory breathing state.

Purpose: Our work aims to add a new perspective to these previous assessments by locally comparing dark-field lung information between different respiratory states to explore new ways of functional lung imaging based on dark-field chest radiography.

Methods: We use suitable deformable image registration methods for dark-field chest radiographs to establish a mapping of lung areas in distinct breathing states. After registration, we utilize an inter-frame ratio approach to examine the local dark-field signal changes and evaluate the gradient of the craniocaudal axis projections and mean lung field values to draw a quantitative comparison to standard chest radiographs and assess the relationship with the respiratory capacity.

Results: Considering full inspiration and expiration scans from a clinical chronic obstructive pulmonary disease study, the registration framework allows to establish an accurate spatial correspondence (Median Dice score 0.95/0.94, mean surface distance 3.71/3.52 mm, and target registration error 6.10 mm) between dark-field chest radiographs in different respiratory states and thus to perform a local signal change analysis. Compared to the utilization of standard chest radiographs, the presented approach benefits from the absence of bone and soft-tissue structures in the dark-field images, which move differently during respiration than the lung tissue. Our quantitative evaluation of the inter-frame ratios demonstrates evidence of craniocaudal gradient-sensitivity advantages concerning the relative vital lung capacity of the study participants in the dark-field images (Spearman correlation coefficients: , and , compared to the attenuation image-based gradient correlations , and , ). Moreover, our alternative lung field analysis approach provides insights into the distinct behavior of the dark-field signal changes with the breathing capacity, which are in good agreement with the expected lung volume changes in the respective lung regions. In quantitative terms, this is reflected in a weak Spearman correlation ( , ) of the mean dark-field signal ratio within the upper lung region, but strong correlations within the middle ( , ) and lower ( , ) lung region.

Conclusions: Our regional characterization of lung dark-field signal changes between the breathing states via deformable image registration provides a proof-of-principle that dynamic radiography-based lung function assessment approaches may benefit from considering registered dark-field images in addition to standard plain chest radiographs. This opens up new options for low-dose and rapid lung ventilation assessment via dark-field chest radiography that has the potential to improve lung diagnostics considerably.