Inter-observer Variation in CT Diagnosis of Solitary Pulmonary Mass

WANG Qiuping,LI Yan,NIU Gang,ZHANG Yun,SUN Meng,QIANG Yongqian,YAN Rui,GUO Youmin
DOI: https://doi.org/10.3969/j.issn.1002-1671.2011.08.007
2008-01-01
Abstract:Objective To assess inter-observer variation in CT evaluation of solitary pulmonary mass. Methods Forty-nine cases with pathologically proved solitary pulmonary masses were independently observed by three radiologists with different working experiences on CT image. Evaluation index included demarcation, boundary contour, lobulation, spiculate protuberance, sentus sign, density, vacuole sign, air-bronchogram, cavitas, calcification, vessel convergence, pleural indentation sign and the nature of masses (malignant or benign). The results were analyzed with Kappa test (K value). Results The consistency of recognization of masses’ nature was moderate between every two observers [observer 1 vs 2, consistency rate 79.59% (39/49, K=0.435, P=0.001), observer 2 vs 3, consistency rate, 85.71% (42/49, K=0.644, P=0.000), observer 1 vs 3, consistency rate 81.63% (40/49, K=0.422, P=0.002). Inter-observer consistency for the presence of boundary contour was moderate, so as inter-observer agreement for the presence of boundary contour, whereas inter-observer consistency for the presence of spiculate protuberance, sentus sign, density, air-bronchogram, and vessel convergence was low, and for the presence of the other CT signs were different. Conclusion There are differences for evaluation of CT signs of solitary pulmonary masses in three radiologists of different working experiences.
What problem does this paper attempt to address?