Adversarial Deformation Regularization for Training Image Registration Neural Networks.
Yipeng HuEli GibsonNooshin GhavamiEster BonmatiCaroline M. MooreMark EmbertonTom VercauterenJ. Alison NobleDean C. BarrattPublished in: MICCAI (1) (2018)
Keyphrases
- image registration
- neural network
- training process
- training algorithm
- mutual information
- deformation field
- feed forward neural networks
- feedforward neural networks
- pattern recognition
- artificial neural networks
- nonrigid registration
- image pairs
- coarse to fine
- affine transformation
- backpropagation algorithm
- deformation model
- registration framework
- early stopping
- image processing
- stochastic gradient descent
- error back propagation
- multi layer
- medical imaging
- image analysis
- back propagation
- multilayer perceptron
- neural network training
- training set
- image alignment
- registration process
- multi modality
- deformable registration
- genetic algorithm
- free form deformation
- fuzzy logic
- supervised learning
- feature matching
- radial basis function network
- displacement field
- registration accuracy
- registration errors
- semi supervised
- image restoration
- multiresolution
- image matching
- recurrent networks
- prior information
- neural nets
- multi agent
- recurrent neural networks