Handling Varying Number of Segments in Custom CT Scan Dataset for nnUNet

[CGraduateTRM]

Hello everyone,

I'm currently working on creating a custom dataset of CT scans for training with nnUNet. In my dataset, the number of segments (and consequently, the number of labels) varies between different training cases. I'm wondering if it's possible to handle such variability in nnUNet and how to ensure dataset integrity verification without encountering RuntimeErrors during preprocessing.

Specifically, my question is twofold:

Is it feasible to have a custom dataset of CT scans where the number of segments varies between training cases? Are there any specific considerations or limitations I should be aware of in this scenario?

How should I appropriately structure my dataset.json file to accommodate varying numbers of segments while ensuring that dataset integrity can be verified without encountering RuntimeErrors during preprocessing?

Any insights, recommendations, or best practices on how to address this issue would be greatly appreciated.

Best regards,