# DICOM gray level normalization , 3 sigma normalization

**URL:** <https://discourse.slicer.org/t/dicom-gray-level-normalization-3-sigma-normalization/18518>\
**Category:** Radiomics\
**Created:** [July 5, 2021, 2:09pm UTC](https://discourse.slicer.org/t/dicom-gray-level-normalization-3-sigma-normalization/18518 "2021-07-05T14:09:52Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![hamdullaherk](https://avatars.discourse-cdn.com/v4/letter/h/bc79bd/32.png) [@hamdullaherk](https://discourse.slicer.org/u/hamdullaherk)\
**Post date:** [July 5, 2021, 2:09pm UTC](https://discourse.slicer.org/t/dicom-gray-level-normalization-3-sigma-normalization/18518/1 "2021-07-05T14:09:53Z")

</div>

Hi, I am a radiologist who is interests in texture analysis and radomics.  
I am a beginner at 3Dslicer. Firstly I want to ask, how can I use 3sigma normalization technique in DICOM images. I don’t know how can ı do it? can you help me?  
Thanks.

---

<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:** [July 13, 2021, 8:11pm UTC](https://discourse.slicer.org/t/dicom-gray-level-normalization-3-sigma-normalization/18518/2 "2021-07-13T20:11:21Z")

</div>

Some normalization can be done in pyradiomics as part of that analysis:

> <https://github.com/AIM-Harvard/pyradiomics/pull/209>
>
> Optionally normalize gray value intensities before any resampling or filters are… applied. Image is normalized on all gray values (not just those in the ROI). Disabled by default. Controlled by 3 new params (\`normalize\`, \`normalizeScale\` and \`removeOutliers\`).

Or you could use numpy operations on the pixel data to implement the statistics and normalization in a few lines.

[https://slicer.readthedocs.io/en/latest/developer\_guide/script\_repository.html#modify-voxels-in-a-volume](https://slicer.readthedocs.io/en/latest/developer_guide/script_repository.html#modify-voxels-in-a-volume)