A cycle generative adversarial network for improving the quality of four-dimensional cone-beam computed tomography images

Table 1 Structures of the generator and discriminator, and gradient optimization conditions

Optimizer	Adam gradient descent method
Minibatch size	1
Initial learning rate	0.0002
Epochs	100
Encoder depth	Generator: 3, discriminator: 4
Convolution filter	Generator: 32, discriminator: 128
Residual block	6

ISSN: 1748-717X