# Discordant shape features

**URL:** <https://discourse.slicer.org/t/discordant-shape-features/9229>\
**Category:** Support\
**Tags:** segmentation, radiomics\
**Created:** [November 20, 2019, 1:12pm UTC](https://discourse.slicer.org/t/discordant-shape-features/9229 "2019-11-20T13:12:05Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![MichaelMoran96](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/michaelmoran96/32/4261_2.png) [@MichaelMoran96](https://discourse.slicer.org/u/MichaelMoran96)\
**Post date:** [November 20, 2019, 1:12pm UTC](https://discourse.slicer.org/t/discordant-shape-features/9229/1 "2019-11-20T13:12:05Z")

</div>

Hi,

I am testing pyradiomic feature repeatability of all seven feature classes(Shape, First-order and 5 Texture features) between different CT images. The phantom i use is a CIRS electron density phantom with 16 solid inserts, therefore, I compare the feature value extracted from insert #1-16 of image 1 with the features extracted from corresponding inserts of image 2

I am using an identical method for segmentation ie creating a 13mm radius cylindrical volume 15mm in length(5x3mm slice) for all inserts in my multi-electron density phantom. The issue i have is no shape feature reaches stability which i have specified at 85% CCC.

I am unsure why shape features are discordant, any help would be greatly appreciated

best wishes,  
Michael