Table 6.
Two-class overlap coefficient of ventilation defect (VD) and ventilated volume (VV) of RVent and JVent with CT as gold standard for different quadrants and registration algorithms.
| ANTs | F-REG | M-REG | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Description | RVent* | JVent | RVent* vs. JVent | RVent* | JVent | RVent* vs. JVent | RVent* | JVent | RVent* vs. JVent |
| UR | 0.72 (0.59–0.89) | 0.62 (0.44–0.77) | 4.54e−07 (****) | 0.68 (0.54–0.86) | 0.57 (0.44–0.71) | 3.21e−08 (****) | 0.46 (0.39–0.54) | 0.37 (0.29–0.53) | 2.81e−10 (****) |
| UL | 0.73 (0.54–0.90) | 0.53 (0.34–0.77) | 2.49e−10 (****) | 0.70 (0.52–0.87) | 0.51 (0.37–0.64) | 3.36e−12 (****) | 0.43 (0.32–0.53) | 0.32 (0.23–0.47) | 2.14e−12 (****) |
| LR | 0.60 (0.52–0.76) | 0.56 (0.42–0.76) | 5.72e−04 (***) | 0.59 (0.51–0.71) | 0.53 (0.42–0.70) | 6.04e−06 (****) | 0.49 (0.41–0.60) | 0.44 (0.35–0.55) | 8.39e−09 (****) |
| LL | 0.58 (0.48–0.74) | 0.53 (0.41–0.83) | 6.97e−02 (n.s.) | 0.55 (0.46–0.68) | 0.50 (0.41–0.76) | 3.69e−03 (**) | 0.47 (0.36–0.56) | 0.42 (0.30–0.53) | 4.15e−06 (****) |
| Omnibus quadrants | 1.19e−11 (****) | 1.23e−01 (n.s.) | — | 1.77e−07 (****) | 2.74e−02 (*) | — | 4.22e−07 (****) | 1.10e−16 (****) | — |
| UR vs. LR | 9.43e−09 (****) | − (−) | — | 2.26e−06 (****) | 3.04e−01 (n.s.) | — | 2.75e−02 (*) | 4.93e−05 (****) | — |
| UL vs. LL | 2.49e−06 (****) | − (−) | — | 5.74e−06 (****) | 3.44e−02 (*) | — | 8.11e−03 (**) | 4.75e−07 (****) | — |
Especially for UR and UL, RVent showed significantly higher overlaps. p ≤ 0.05 (*), p ≤ 0.01 (**), p ≤ 0.001 (***), p ≤ 0.0001 (****). Please note that the numbers displayed in this table are rounded to two decimal places. As a result, different values may appear identical despite minor differences in their actual values. ANTs: advanced normalization tools, F-Reg: Forsberg registration, M-Reg: Matlab registration, RVent*: filtered regional ventilation, JVent: Jacobian determinant ventilation, UR: upper right, UL: upper left, LR: lower right, LL: lower left, VD: ventilation defect.