Table 3.
Comparison | Volume | Intraclass correlation | Bias | Std. deviation | Lower LOA | Upper LOA |
---|---|---|---|---|---|---|
Diagnostic CT segmentation vs. cuboid approximation | Primary (ml) | 0.60 | +15.9 | 26.1 | −35.3 | +67.1 |
Primary (%) | - | +54.1 | 22.5 | +4.1 | +104.1 | |
Largest LN (ml) | 0.79 | +2.9 | 17.4 | −31.2 | +37.0 | |
Largest LN (%) | - | +37.7 | 40.0 | −40.7 | +116.1 | |
Diagnostic CT segmentation vs. ellipsoid approximation | Primary (ml) | 0.88 | −0.7 | 10.0 | −20.3 | +18.9 |
Primary (%) | - | −8.0 | 27.8 | −64.0 | +48.0 | |
Largest LN (ml) | 0.82 | −13.5 | 31.6 | −75.4 | +48.4 | |
Largest LN (%) | - | −18.0 | 28.3 | −72.0 | +54.0 | |
Diagnostic CT segmentation vs. radiotherapy-planning CT segmentation | Primary (ml) | 0.74 | +5.7 | 16.6 | −26.9 | +38.2 |
Primary (%) | - | +28.1 | 56.0 | −81.7 | +137.9 | |
Largest LN (ml) | 0.96 | +0.2 | 8.2 | −15.9 | +16.2 | |
Largest LN (%) | - | +12.5 | 66.4 | −117.6 | +142.6 |
Due to proportional bias of the raw volume data, only percent difference values were applicable over a wide range of tumor and LN sizes