Skip to main content

Table 1 Detailed information of input images before training

From: A deep learning-based self-adapting ensemble method for segmentation in gynecological brachytherapy

 

2D

3D-fullres

3D-lowres

Median image size

512 × 512

63 × 512 × 512

63 × 354 × 354

Median target spacing

0.75 × 0.75

2.5 × 0.75 × 0.75

2.5 × 1.0838 × 1.0838

Patch size

512 × 512

28 × 256 × 256

40 × 224 × 224

Batch size

12

2

2

Back to article page

Radiation Oncology

ISSN: 1748-717X

Contact us