Comparison of ultrasonography and computed tomography for diagnosing diffuse thyroid disease: a multicenter study

Abstract

To compare the diagnostic performance of ultrasonography (US) and computed tomography (CT) for diagnosing incidentally detected diffuse thyroid disease (DTD) in patients who underwent thyroid surgery using multicenter data. Between July and December 2016, a total of 177 patients who underwent preoperative thyroid US and neck CT, and subsequent thyroid surgery at 4 participating institutions, were reviewed. US and CT images in each case were retrospectively reviewed by a radiologist at each institution, and classified into one of the following four categories based on US and CT features: no DTD; indeterminate; suspicious for DTD; and DTD. The diagnostic accuracy of US and CT were calculated at each institution by comparison with histopathological results. Respective US and CT classifications in the 177 patients were no DTD in 75 and 71, indeterminate in 46 and 34, suspicious for DTD in 28 and 31, and DTD in 28 and 41. Among the histopathological results, 113 patients had normal thyroid parenchyma, 23 had Hashimoto thyroiditis, 36 had non-Hashimoto lymphocytic thyroiditis, and 5 had diffuse hyperplasia. The presence of >= 2 US and CT features of DTD, which was classified as suspicious for DTD or DTD, had the largest area under the receiver operating characteristic curve (0.866 and 0.893, respectively), with sensitivity and specificity of 71.9 and 91.2% in US, and 84.4 and 84.1% in CT, respectively. However, there was no statistically significant difference between readers' experience and their diagnostic performance. US and CT imaging may be helpful for detecting incidental DTD.