# Radiomics stability analysis

**URL:** <https://discourse.slicer.org/t/radiomics-stability-analysis/22682>\
**Category:** Radiomics\
**Tags:** radiomics, pyradiomics\
**Created:** [March 25, 2022, 2:34pm UTC](https://discourse.slicer.org/t/radiomics-stability-analysis/22682 "2022-03-25T14:34:12Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![JonasB](https://avatars.discourse-cdn.com/v4/letter/j/8e7dd6/32.png) [@JonasB](https://discourse.slicer.org/u/JonasB)\
**Post date:** [March 25, 2022, 2:34pm UTC](https://discourse.slicer.org/t/radiomics-stability-analysis/22682/1 "2022-03-25T14:34:12Z")

</div>

Hi,

I am currently searching for a good method to test for radiomics feature stability.  
Unfortionetly, I do not have the opportunity to do a test-retesting as I am working with retrospective data.  
Is there any good method to do it and is the interclass correlation coefficient (ICC) the best choice?

Best,  
Jonas

---

<div class="post-metadata">

**Author:** ![AndreasL](https://avatars.discourse-cdn.com/v4/letter/a/aeb1de/32.png) [@AndreasL](https://discourse.slicer.org/u/AndreasL)\
**Post date:** [March 19, 2023, 7:15pm UTC](https://discourse.slicer.org/t/radiomics-stability-analysis/22682/2 "2023-03-19T19:15:41Z")

</div>

Hi Jonas,

did you figure it out?

Andreas