Closed Li-rr closed 2 years ago
Just going through some old / open issues. I thought about it but it has become too much of a hassle to pull all the parts together sorry. The training code itself is very simple, essentially just a 2D segmentation network with a Binary Cross Entropy loss. Voxelising the dataset and making sure it is all aligned is the hard (and computationally expensive) part.
Very cool paper. Do you have any plans to release it?