Skip to main content
. Author manuscript; available in PMC: 2021 Mar 1.
Published in final edited form as: Med Phys. 2019 Dec 29;47(3):837–849. doi: 10.1002/mp.13955

Figure 4:

Figure 4:

Deep learning models used in the study. The same U-net architecture is utilized in each comparison of MSE, MSE+ADV, MSE+DVH, and MSE+DVH+ADV models. The same discriminator architecture is utilized for training the MSE+ADV and MSE+DVH+ADV models. Black numbers to the left of the feature blocks represent the current data shape. Red numbers above the feature blocks represents the number of features.