Exported ROI Collections Mask Segmentation

[Jun Xian Tan]

Hello, I have been experimenting around annotating Mask for CT scans in the viewer. I’m using version 3.5.1, and the annotation process seems fine so far, until the exportation part.

‌

I have done some annotations, and I have labelled them different names:

‌

Exporting this masked-based ROI Collection works without any error:

‌

However, when i downloaded the exported mask segmentation, tried to view it on 3D Slicer/ITK-Snap, it doesn’t separate it by label, but it’s only a binary segmentation.

Downloading the exported seg:

‌

Viewing the exported mask on ITK-Snap:
It seems like the segmentation is separated by label as we can see z=22 (22 slices) which matches the 14 (ich label) + 8 (ich2 label)

Viewing it on 3D Slicer, setting to “Labels” still showed the 2 different labels as same color:

Is this the intended working way? Or is there a way I can export the SEG with different labels/separate the labels out

‌

[Jun Xian Tan]

Attached the CT scan with the exported annotation here

[Jun Xian Tan]

After downloading the SEG, i deleted the ROI Collection from xnat:

And then uploaded the downloaded SEG into xnat:

And it seems successful upon importing in the viewer:

‌

[Jun Xian Tan]

Hello again,

So after reading up on DICOM Seg, I found out that this exported file has to be converted. I did so by following this: https://razorx89.github.io/pydicom-seg/guides/read.html
After the conversion, reading the new .nrrd file on ITK-Snap showed the different labels:

I think this issue can be closed

Edit: Would it be possible to include the empty slices in the export process as well?