Purpose: To test the hypothesis that the joint distribution of x-ray transmission and dark-field signals obtained with a compact cone-beam preclinical scanner with a polychromatic source can be used to diagnose pulmonary emphysema in ex vivo murine lungs.
Materials and methods: The animal care committee approved this study. Three excised murine lungs with pulmonary emphysema and three excised murine control lungs were imaged ex vivo by using a grating-based micro-computed tomographic (CT) scanner. To evaluate the diagnostic value, the natural logarithm of relative transmission and the natural logarithm of dark-field scatter signal were plotted on a per-pixel basis on a scatterplot. Probability density function was fit to the joint distribution by using principle component analysis. An emphysema map was calculated based on the fitted probability density function.
Results: The two-dimensional scatterplot showed a characteristic difference between control and emphysematous lungs. Control lungs had lower average median logarithmic transmission (-0.29 vs -0.18, P = .1) and lower average dark-field signal (-0.54 vs -0.37, P = .1) than emphysematous lungs. The angle to the vertical axis of the fitted regions also varied significantly (7.8° for control lungs vs 15.9° for emphysematous lungs). The calculated emphysema distribution map showed good agreement with histologic findings.
Conclusion: X-ray dark-field scatter images of murine lungs obtained with a preclinical scanner can be used in the diagnosis of pulmonary emphysema.
Supplemental material: http://radiology.rsna.org/lookup/suppl/doi:10.1148/radiol.13122413/-/DC1.
RSNA, 2013