Table 2 Comparison of the BraTS 2020 segmentation results for the ablation experiment of the dual transformer encoder.

From: DTASUnet: a local and global dual transformer with the attention supervision U-network for brain tumor segmentation

Method

Dice

Sensitivity

ET

WT

TC

Mean

ET

WT

TC

Mean

DTASUnet with 3D fix local transformer

0.769

0.909

0.855

0.844

0.764

0.930

0.885

0.860

DTASUnet with 3D shift local transformer

0.808

0.924

0.876

0.869

0.805

0.925

0.874

0.868

DTASUnet with 3D global transformer

0.770

0.919

0.869

0.853

0.770

0.927

0.887

0.861

DTASUnet with 3D dual transformer

0.836

0.925

0.885

0.882

0.842

0.920

0.878

0.880

  1. Bold results denote the best score.
Back to article page