Volumetric Medical Image Registration on Brain MRI (test)
72.6DiceViT-V-Net
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ViT-V-NetHardware=GPU, Optimizer=ADAM, Loss function=MSE, Regularizer=Diffusion, Regularization parameter (lambda)=0.02, ViT patch size (P)=8, ViT latent vector size (D)=2522021.04 | 72.6 | 38.1 | 0.002 | |
| NiftyRegHardware=CPU, Cost function=SSD, Regularizer=Bending energy, Regularization parameter (lambda)=0.00022021.04 | 71.3 | 22.5 | 113 | |
| VoxelMorph-2Hardware=GPU, Optimizer=ADAM, Loss function=MSE, Regularizer=Diffusion, Regularization parameter (lambda)=0.022021.04 | 71.1 | 41.4 | 0.002 | |
| VoxelMorph-1Hardware=GPU, Optimizer=ADAM, Loss function=MSE, Regularizer=Diffusion, Regularization parameter (lambda)=0.022021.04 | 70.7 | 37.5 | 0.002 | |
| SyNHardware=CPU, Cost function=MSQ, Regularizer=Gaussian, Regularization parameter (lambda)=32021.04 | 68.8 | 11.8 | 15.257 |