# Scalar-Weighted PCA

**URL:** <https://discourse.slicer.org/t/scalar-weighted-pca/40608>\
**Category:** Support\
**Tags:** segmentation, dicom, quantitative, pca\
**Created:** [December 10, 2024, 3:45pm UTC](https://discourse.slicer.org/t/scalar-weighted-pca/40608 "2024-12-10T15:45:43Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![francois](https://avatars.discourse-cdn.com/v4/letter/f/77aa72/32.png) [@francois](https://discourse.slicer.org/u/francois)\
**Post date:** [December 10, 2024, 3:45pm UTC](https://discourse.slicer.org/t/scalar-weighted-pca/40608/1 "2024-12-10T15:45:43Z")

</div>

Hello,  
I am looking at Foot CT scans. I have segmented each bone. I would like to compute the PCA axes of each bone, WEIGHTED with the scalar value of the voxels contained/overlapping each bone model. So the value for each voxel would be [HU\*(x,y,z)]. Is there any way to do that using the quantifying module or some other way ? Label map statistics seem to only take into account spatial data and scalar statistics only give min, max, mean, median, sd values for the considered segment.  
Thank you !