# Some loss for convert the stl model to the editable binary labelmap

**URL:** <https://discourse.slicer.org/t/some-loss-for-convert-the-stl-model-to-the-editable-binary-labelmap/17987>\
**Category:** Support\
**Tags:** segmentation\
**Created:** [June 7, 2021, 2:46am UTC](https://discourse.slicer.org/t/some-loss-for-convert-the-stl-model-to-the-editable-binary-labelmap/17987 "2021-06-07T02:46:37Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![xiang\_luo](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/xiang_luo/32/9305_2.png) [@xiang\_luo](https://discourse.slicer.org/u/xiang_luo)\
**Post date:** [June 7, 2021, 2:46am UTC](https://discourse.slicer.org/t/some-loss-for-convert-the-stl-model-to-the-editable-binary-labelmap/17987/1 "2021-06-07T02:46:37Z")

</div>

When i import a .stl file by the import data module, there is no anything unusual for the model visualization. But when i want to edit it for some fixed, i convert it to the binary label map and segmentation node, then set them as “master volume” and “segmentation node” separately in the segmentation node. And i find there some thin struture vessel disappear when i use the “Island-Keep largest island”. I don’t know why this happend, can you fix this?

---

<div class="post-metadata">

**Author:** ![lassoan](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/lassoan/32/13_2.png) [@lassoan](https://discourse.slicer.org/u/lassoan)\
**Post date:** [June 7, 2021, 3:30am UTC](https://discourse.slicer.org/t/some-loss-for-convert-the-stl-model-to-the-editable-binary-labelmap/17987/2 "2021-06-07T03:30:47Z")

</div>

When a segmentation is edited, then it is converted to binary labelmap representation. By default this labelmap is converted to about 16 million voxels, which may or may not be enough to preserve all details that you are interested in. You can [increase the binary labelmap resolution](https://slicer.readthedocs.io/en/latest/user_guide/modules/segmenteditor.html#segmentation-is-not-accurate-enough) to address this.