摘要
Objective：To research on consistency of identifying solitary pulmonary masses with CT. Methods：Three observers with different working backgrounds in imaging diagnosis individually interpreted the same group images of solitary pulmonary mass, by 12 indexes of objective signs. The differences in interpretation resulted in ante- and post-interpretations were assessed by the x^2 test. The agreement of two interpretations from the same observer was confirmed with the kappa test. A double-blind method was adopted for analysis. Results：The agreement rates of ante- and post-interpreting from the three observers were respectively 82.65%（486/588） 80.27%（472/ 588） and 84.86% （499/588） while their interpreting results were generally accordant without significant difference（x^2 = 4.975, df= 2, P= 0.083） however there was difference between the observer 2 and observer 3（x^2 = 4.875, df= 1, P = 0.027）. There were five indexes with k 〉 0.40 of ante- and post-interpreting results of the three observers, including clarity of nodule borderline, presence of sentus, uniformity of density, existence of cavity and calcification in pathological region, among them, the agreement rate of interpreting borderline and cavity was higher（k 〉 0.07）; the blood vessel convergence poorer（0 〈 k ≤ 0.40）; the other six CT signs of interpretation were slightly different. Conclusion：The ability to identify solitary pulmonary mass was inconsistent, and needs to be improved further.
Objective：To research on consistency of identifying solitary pulmonary masses with CT. Methods：Three observers with different working backgrounds in imaging diagnosis individually interpreted the same group images of solitary pulmonary mass, by 12 indexes of objective signs. The differences in interpretation resulted in ante- and post-interpretations were assessed by the x^2 test. The agreement of two interpretations from the same observer was confirmed with the kappa test. A double-blind method was adopted for analysis. Results：The agreement rates of ante- and post-interpreting from the three observers were respectively 82.65%（486/588） 80.27%（472/ 588） and 84.86% （499/588） while their interpreting results were generally accordant without significant difference（x^2 = 4.975, df= 2, P= 0.083） however there was difference between the observer 2 and observer 3（x^2 = 4.875, df= 1, P = 0.027）. There were five indexes with k 〉 0.40 of ante- and post-interpreting results of the three observers, including clarity of nodule borderline, presence of sentus, uniformity of density, existence of cavity and calcification in pathological region, among them, the agreement rate of interpreting borderline and cavity was higher（k 〉 0.07）; the blood vessel convergence poorer（0 〈 k ≤ 0.40）; the other six CT signs of interpretation were slightly different. Conclusion：The ability to identify solitary pulmonary mass was inconsistent, and needs to be improved further.
