Comparison of volumetric and bidimensional measurement of the thymic gland to determine interobserver variability

Abstract

Aim: Thymic measurements should be interpreted relative to normal measurement values according to age groups and gender. Thymic gland measurement may vary between radiologists. In our study we wanted to determine the interobserver variability of thymic size measurements on CT by comparing volumetric calculations to commonly used bidimensional measurements.
 Material and Method: The chest CT scans of 156 patients who had any oncological problems or disorders influencing the thymic size were retrospectively evaluated. The shape (quadrilateral or triangular), anteroposterior (AP) and transverse (TR) diameters, as well as the thickness of each lobe and thymus volume were measured independently by two radiologists. Interobserver variability was determined according to these parameters by Pearson correlation test.
 Results: The shape of the thymic gland was triangular in 72% (n:113) and quadrilateral in 28% (n: 43) of all subjects. According to the results of the correlation test, a moderate relationship in terms of AP length, a weak relationship in terms of TR length, a negligible relationship in terms of right lobe thickness and a weak relationship in terms of left lobe thickness were obtained, but a very strong relationship was found between the two radiologists for volumetric measurements.
 Conclusions: Volumetric measurements can be used as a basis for thymic imaging, reducing the differences among radiologists, and unnecessary and advanced examinations can be avoided.