Comparison of four software packages for CT lung volumetry in healthy individuals

Objectives To compare CT lung volumetry (CTLV) measurements provided by different software packages, and to provide normative data for lung densitometric measurements in healthy individuals. Methods This retrospective study included 51 chest CTs of 17 volunteers (eight men and nine women; mean age, 30 ± 6 years), who underwent spirometrically monitored CT at total lung capacity (TLC), functional residual capacity (FRC), and mean inspiratory capacity (MIC). Volumetric differences assessed by four commercial software packages were compared with analysis of variance (ANOVA) for repeated measurements and benchmarked against the threshold for acceptable variability between spirometric measurements. Mean lung density (MLD) and parenchymal heterogeneity (MLD-SD) were also compared with ANOVA. Results Volumetric differences ranged from 12 to 213 ml (0.20 % to 6.45 %). Although 16/18 comparisons (among four software packages at TLC, MIC, and FRC) were statistically significant ( P < 0.001 to P = 0.004), only 3/18 comparisons, one at MIC and two at FRC, exceeded the spirometry variability threshold. MLD and MLD-SD significantly increased with decreasing volumes, and were significantly larger in lower compared to upper lobes ( P < 0.001). Conclusions Lung volumetric differences provided by different software packages are small. These differences should not be interpreted based on statistical significance alone, but together with absolute volumetric differences. Key Points • Volumetric differences, assessed by different CTLV software, are small but statistically significant . • Volumetric differences are smaller at TLC than at MIC and FRC . • Volumetric differences rarely exceed spirometric repeatability thresholds at MIC and FRC . • Differences between CTLV measurements should be interpreted based on comparison of absolute differences . • MLD increases with decreasing volumes, and is larger in lower compared to upper lobes .