Skip to main content

Table 1 Multiple comparison of experimental result indicators of 6 networks (\(\begin{gathered} \overline{x} \pm s \hfill \\ \hfill \\ \end{gathered}\))

From: Automatic segmentation of the gross target volume in radiotherapy for lung cancer using transresSEUnet 2.5D Network

Metric

Model(\(\overline{X} \pm S\))

P

95% Confidence interval

Lower bound

Upper bound

DSC (%)

TransResSEUnet2.5D (84.08 ± 0.04)

Unet2D (77.07 ± 0.09)

0.090

− 0.006

0.146

Unet2.5D (81.51 ± 0.06)

0.910

− 0.032

0.083

Unet3D (74.53 ± 0.17)

0.358

− 0.419

0.233

ResSEUnet3D (80.56 ± 0.06)

1.000

− 0.042

0.064

ResSEUnet2.5D (82.97 ± 0.06)

0.456

− 0.018

0.088

HD95 (mm)

TransResSEUnet2.5D (8.11 ± 3.43)

Unet2D (15.25 ± 9.04)

0.050

− 14.278

0.002

Unet2.5D (9.69 ± 6.30)

0.996

− 6.797

3.625

Unet3D (13.91 ± 13.47)

0.662

− 16.179

4.594

ResSEUnet3D (11.14 ± 5.43)

1.000

− 4.932

3.550

ResSEUnet2.5D (8.80 ± 4.81)

0.486

− 7.658

1.612

Average prediction time of single series (s)

TransResSEUnet2.5D (6.50 ± 1.31)

Unet2D (6.30 ± 2.29)

0.944

− 1.377

0.807

Unet2.5D (3.69 ± 0.72)

0.000

1.644

3.889

Unet3D (3.51 ± 0.58)

0.000

1.853

4.035

ResSEUnet3D (4.69 ± 0.96)

0.301

− 0.318

2.191

ResSEUnet2.5D (5.52 ± 1.10)

0.001

0.557

2.9574