In the liver and tumor segmentation model, it's possible to approximate the Brain Segmentation workflow with RepeatChannelD to enable usage of ConvertToMultiChannelBasedOnBratsClassesd. However this results in greater, unnecessary memory usage.
Adding a DeleteChannelD transform would allow the duplicated channels to be pared down before feeding into the network.
In the liver and tumor segmentation model, it's possible to approximate the Brain Segmentation workflow with RepeatChannelD to enable usage of ConvertToMultiChannelBasedOnBratsClassesd. However this results in greater, unnecessary memory usage.
Adding a DeleteChannelD transform would allow the duplicated channels to be pared down before feeding into the network.