Segmentations using HU when Thresholding is not accurate enough

Screen Shot 2022-06-16 at 10.29.29 AM
Screen Shot 2022-06-16 at 10.19.03 AM
Screen Shot 2022-06-16 at 10.19.16 AM
Screen Shot 2022-06-16 at 10.29.40 AM

Hi All,

I have a “heat map” DICOM here that I brought into slicer and I need each different color (see images) to be a different segmentation. I cannot use the threshold effect though as each color corresponds to a single HU (ie. 1,2,3,4,5 etc. see images attached in bottom left with green circle).

Is the best way to do this to break it down into a numpy array and then assign a segmentation to a particular value? Is that even possible? Does anyone else have any ideas on how to solve this problem?

Thanks,
Zach

I would try using a module to convert it to a gray scale image.
It is called RGB to scalar or something like it.
Then you’d be able to threshold I think

You can do this by loading this image as a segmentation as described here. It is unnecessary to load each label as a separate segmentation, it is sufficient to load it as a separate segment.

However, you can display much nicer colormaps by loading the image as a scalar volume and then choosing an appropriate color lookup table in volumes module. See for example how dose maps are rendered in SlicerRT extension.