Skip to main content

Table 2 The mean absolute error results of the body for different groups in seven fold cross-validation

From: Dose prediction for cervical cancer VMAT patients with a full-scale 3D-cGAN-based model and the comparison of different input data on the prediction results

Groups of cross-validation

Model-A

Model-B

Model-C

MAE-A (%)

P1

MAE-B (%)

P2

MAE-C (%)

P3

1

0.9 ± 0.4

0.102

1.3 ± 0.3

0.389

1.8 ± 0.3

0.221

2

1.1 ± 0.6

0.930

1.5 ± 0.3

0.405

2.2 ± 0.4

0.089

3

1.0 ± 0.3

0.814

1.2 ± 0.2

0.097

2.3 ± 0.4

0.080

4

1.0 ± 0.4

0.858

1.4 ± 0.3

0.762

2.0 ± 0.2

0.035

5

1.3 ± 0.4

0.791

1.5 ± 0.2

0.544

1.9 ± 0.6

0.216

6

1.4 ± 0.6

0.067

1.6 ± 0.6

0.085

2.3 ± 0.5

0.073

7

0.9 ± 0.2

0.083

1.3 ± 0.3

0.230

1.8 ± 0.3

0.229

  1. Annotate: MAE-A, B, C represent the mean absolute error of each cross-validation group of Model-A, B, C, respectively. P1, P2, P3, represent the paired t-test results of each cross-validation group and the test group of Model-A, B, C, respectively