Question About Flexible Slice Implementation in Your 2.5D Segmentation Code

[Yanqs2000]

I am currently conducting research using the code from your paper “A flexible 2.5D medical image segmentation approach with in-slice and cross-slice attention”. In your paper, you mention that the model can accommodate an arbitrary number of slices. However, while reviewing the code, I only found an implementation for three slices.

Would you be able to provide a version of the code that supports an arbitrary number of slices, or offer guidance on how to adapt the current implementation for this purpose?

[qinxu888]

你说的这个我做过了，改跨切片注意力结构和卷积