Skip to main content

Table 1 Hyper-parameters for training deep learning-based synthetic contrast-enhanced computed tomography generation model

From: Synthetic contrast-enhanced computed tomography generation using a deep convolutional neural network for cardiac substructure delineation in breast cancer radiation therapy: a feasibility study

Parameter

Value

No. of parameters

\(G\): 5.4 M/\(D\): 1.6 M

Batch size

4

Loss function

Adversarial + L1 loss

Optimizers

\(G\): Adam/\(D\): SGD

Starting learning rate

\(G\): 0.0002/\(D\): 0.00002

Number of epochs

200

  1. G, generator; D, discriminator; SGD, stochastic gradient descent
Back to article page

Radiation Oncology

ISSN: 1748-717X

Contact us