Fig 1.
Example of a case excluded due to very diffuse contrast enhancing properties.
Fig 2.
Examples of user-interface for segmentation in each software package.
a BrainVoyager, b 3D Slicer, c ITK-Snap.
Fig 3.
Mean volumes with 95% CI for the two raters in the different software solutions.
a Segmentations by A.L.S, b segmentations by E.H.F.
Fig 4.
Difference in tumor volumes for second segmentations between investigators using each software plotted against mean tumor volume of each segmentation (inter-rater agreement). Whole line represents mean difference, stapled lines represent 95% limits of agreement. a BrainVoyager, b 3D Slicer, c ITK-Snap
Fig 5.
Difference in tumor volumes for combined mean of second segmentations for investigators between software packages plotted against mean tumor volume of each segmentation. Whole line represents mean difference, stapled lines represent 95% limits of agreement. a BrainVoyager and 3D Slicer, b BrainVoyager and ITK-Snap, c ITK-Snap and 3D Slicer
Table 1.
Descriptive statistics for tumor volume segmentations in mL for each investigator (N = 18).
Table 2.
Intra-rater, Inter-rater and Between-software agreement reported as difference of means (mL) with 95% LoA.
Table 3.
Inter-rater Dice’s similarity coefficients (DSC) and Hausdorff distance in millimeters (HD) (N = 18).
Table 4.
Between software Dice’s similarity coefficients (DSC) and Hausdorff distance in millimeters (HD), for each rater (N = 18).