Fig. 2
From: Artifact-free fat-water separation in Dixon MRI using deep learning
Proposed 3D cGAN architecture with IP and OP signals as input and separated \(\hat{F}\) and \(\hat{W}\) signals as output. Filter numbers for each convolutional block are provided. The contracting path of the generator is in orange and the expansive path is in blue. The discriminator is shown in green, the different shades indicate the change in stride length