Bone images from dual-energy subtraction chest radiography in the detection of rib fractures

Objective: To assess the sensitivity and image quality of chest radiography (CXR) with or without dual-energy subtracted (ES) bone images in the detection of rib fractures. Materials and methods: In this retrospective study, 39 patients with 204 rib fractures and 24 subjects with no fractures were examined with a single exposure dual-energy subtraction digital radiography system. Three blinded readers first evaluated the non-subtracted posteroanterior and lateral chest radiographs alone, and 3 months later they evaluated the non-subtracted images together with the subtracted posteroanterior bone images. The locations of rib fractures were registered with confidence levels on a 3-grade scale. Image quality was rated on a 5-point scale. Marks by readers were compared with fracture localizations in CT as a standard of reference. Results: The sensivity for fracture detection using both methods was very similar (34.3% with standard CXR and 33.5% with ES-CXR, p = 0.92). At the patient level, both sensitivity (71.8%) and specificity (92.9%) with or without ES were identical. Diagnostic confidence was not significantly different (2.61 with CXR and 2.75 with ES-CXR, p = 0.063). Image quality with ES was rated higher than that on standard CXR (4.08 vs. 3.74, p < 0.001). Conclusions: Despite a better image quality, adding ES bone images to standard radiographs of the chest does not provide better sensitivity or improved diagnostic confidence in the detection of rib fractures. (C) 2010 Elsevier Ireland Ltd. All rights reserved.