Objective: The purpose of this study was to compare tomosynthesis with radiography for the detection of bone erosions of the foot in patients with established rheumatoid arthritis (RA) using MDCT as a reference standard.
Subjects and methods: Eighteen consecutive patients with established RA were included. Each patient underwent radiography, tomosynthesis, and CT examinations of the feet on the same day. Two radiologists independently determined the number of bone erosions and the Sharp-van der Heijde score with each of the three imaging modalities.
Results: On a total of 216 joints from 18 patients, 216 bone erosions were detected on CT, 215 on tomosynthesis, and 181 with radiography. The mean (± SD) Sharp-van der Heijde score was equivalent for tomosynthesis (18.8 ± 16.8) and CT (19.8 ± 18.5) but was statistically lower for radiography (16.4 ± 18.0) (p = 0.030). The respective overall sensitivity, specificity, accuracy, positive predictive value, and negative predictive value for tomosynthesis were 80%, 75%, 78%, 76%, and 80%, whereas the respective corresponding values for radiography were 66%, 81%, 74%, 77%, and 71%. The radiation burden of tomosynthesis was almost equivalent to that of radiography.
Conclusion: Tomosynthesis has a higher sensitivity than radiography to detect bone erosions of the foot in patients with established RA and imparts an almost equivalent radiation burden.
Keywords: arthritis; foot; image enhancement methods; radiography; rheumatoid; tomosynthesis.