Replies: 1 comment
-
You're correct that nnU-Net typically splits 3D images into patches before processing. If you want to process the full 3D image without splitting, you can adjust the patch size to match the full image dimensions. Here's how: Set Patch Size: In your configuration, set the patch size to the full image dimensions. Example: patch_size = (full_image_depth, full_image_height, full_image_width)
batch_size = 1 # Handle one full image at a time This will allow nnU-Net to process the full 3D image without splitting it into patches. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi,
I'm working with nn-UNet to perform the segmentation of some MRI images. I've noticed (correct me if I'm wrong) that the 3d images get split in patches before being analyzed by the CNN, and for my specific task, I think this is important.
Is there the possibility of blocking such behavior? I would like the full 3d image to get passed to the UNet as it is.
Any help is appreciated,
Thanks.
Giorgio
Beta Was this translation helpful? Give feedback.
All reactions