Table 8 Quantitative results of models trained with different Convolution iterations on the LPBA40 and Mindboggle101 datasets.

From: Hybrid transformer and convolution iteratively optimized pyramid network for brain large deformation image registration

Variant model

LPBA40

Mindboggle101

\(\:\text{D}\text{S}\text{C}\uparrow\:\)

\(\:\text{A}\text{S}\text{S}\text{D}\downarrow\:\)

\(\:\left|{J}_{\varnothing\:}\right|<0\downarrow\:\)

\(\:\text{T}\left(\text{s}\right)\downarrow\:\)

\(\:\text{D}\text{S}\text{C}\uparrow\:\)

\(\:\text{A}\text{S}\text{S}\text{D}\downarrow\:\)

\(\:\left|{J}_{\varnothing\:}\right|<0\downarrow\:\)

\(\:\text{T}\left(\text{s}\right)\downarrow\:\)

baseline

0.732

1.428

< 0.027%

0.376

0.633

0.881

< 0.089%

0.372

1_1iter

0.732

1.427

< 0.0058%

0.381

0.628

0.890

< 0.016%

0.406

2_2iter

0.728

1.429

< 0.0075%

0.385

0.628

0.885

< 0.015%

0.416

1_1_1iter(Ours)

0.737

1.404

< 0.0165%

0.419

0.637

0.878

< 0.0495%

0.427

2_2_2iter

0.730

1.426

< 0.0034%

0.452

0.633

0.879

< 0.077%

0.440

Back to article page