Purpose: To prospectively assess and compare the measurement variability of volume and mass for pure and part-solid ground-glass nodules (GGNs) with solid portions less than or equal to 5 mm by using a commercially available volumetric software program.
Materials and methods: This prospective study was approved by the institutional review board. Written informed consent was obtained. From November 2011 to June 2012, 73 patients (26 men and 47 women) with 94 GGNs (>5 mm and <20 mm; 72 pure and 22 part-solid GGNs) were prospectively enrolled and underwent two consecutive computed tomographic (CT) examinations. Both the volume and mass of GGNs were measured with volumetric software by two radiologists. Intraobserver, interobserver, and interscan variability were analyzed and compared by using the Bland-Altman method and coefficients of variation. The influence of the solid portion of GGNs and GGN size on interscan variability was investigated with multiple linear regression analysis and analysis of variance.
Results: Nodule segmentation was successful in 420 of 438 (95.9%) segmentations. As for volume measurement, interscan variability ranged from -17.3% to 18.5%, while intraobserver and interobserver variability ranged from -7.6% to 8.5% and from -11.7% to 18.1%, respectively. Interscan variability in mass measurement ranged from -17.7% to 18.6%, while intraobserver and interobserver variability ranged from -8.4% to 9.4% and from -17.5% to 11.8%, respectively. In the coefficient of variation comparison, there were no significant differences in volume and mass measurements for intraobserver, interscan, and interobserver variability. Measurement variability of volume and mass was not significantly influenced by the presence of a solid portion, solid portion size, or GGN size.
Conclusion: Mass measurement of GGNs showed measurement variability from -17.7% to 18.6% and may be a useful method in the follow-up of GGNs with solid portions less than or equal to 5 mm.