# STL Extraction from CT scans

**URL:** <https://discourse.slicer.org/t/stl-extraction-from-ct-scans/26950>\
**Category:** Support\
**Tags:** segmentation, dicom, 3d-model\
**Created:** [December 27, 2022, 11:22pm UTC](https://discourse.slicer.org/t/stl-extraction-from-ct-scans/26950 "2022-12-27T23:22:19Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![detroitkiddo](https://avatars.discourse-cdn.com/v4/letter/d/a3d4f5/32.png) [@detroitkiddo](https://discourse.slicer.org/u/detroitkiddo)\
**Post date:** [December 27, 2022, 11:22pm UTC](https://discourse.slicer.org/t/stl-extraction-from-ct-scans/26950/1 "2022-12-27T23:22:19Z")

</div>

Hello all,

I am new to the forums and I am reaching out to ask a question regarding segmentation and 3D STL extractions from CT scans compared to MRIs. So far I am had no issues trying to extract the ventricular system (lateral and third) using MRI DICOM scans. However when I try to do the same with CT scans instead the quality is extremely bad and results in very poor 3D renders of the ventricular system. I am able to get great 3D renders of the skull using CT scans but extremely bad renders for the ventricles.

I wanted to ask if using CT scans for the ventricular system is just not possible and I am better off with MRIs or if perhaps I am using the wrong settings for CT scans. Currently for MRIs I use the segmentation tools with the right threshold to get my lateral and third ventricles and I am able to make perfect 3D models. When I apply the same technique with CTs it is extremely bad even with varied threshold settings.

Thank you all for the feedback!

---

<div class="post-metadata">

**Author:** ![pieper](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/pieper/32/8_2.png) [@pieper](https://discourse.slicer.org/u/pieper)\
**Post date:** [December 27, 2022, 11:51pm UTC](https://discourse.slicer.org/t/stl-extraction-from-ct-scans/26950/2 "2022-12-27T23:51:12Z")

</div>

Yes, CT provides less soft tissue contrast than MR. I would suggest trying [SynthSeg](https://github.com/BBillot/SynthSeg) on your CT. It’s not available directly in Slicer but you can run it outside and easily load the results.