# Longitudinal registration brain MRI

**URL:** <https://discourse.slicer.org/t/longitudinal-registration-brain-mri/13481>\
**Category:** Support\
**Tags:** registration\
**Created:** [September 14, 2020, 12:28pm UTC](https://discourse.slicer.org/t/longitudinal-registration-brain-mri/13481 "2020-09-14T12:28:17Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Marianne](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/marianne/32/7886_2.png) [@Marianne](https://discourse.slicer.org/u/Marianne)\
**Post date:** [September 14, 2020, 12:28pm UTC](https://discourse.slicer.org/t/longitudinal-registration-brain-mri/13481/1 "2020-09-14T12:28:17Z")

</div>

What is the best way to compare segments from MRIs with different geometry (individual scans of the same subject)? Do I have to register all the MRIs in order to compare the segments to each other? What is the easiest way to do this? I have not had luck with affine or landmark registration, there is no change “sticking” to the output volume when I change module…

Context: I have a set of MRI-images with corresponding manual and machine-learning-derived tumor segments, shoving longitudinal tumor developments. I want to register all the MRIs to the first MRI, so the geometry of the segments can be compared in respekt of visualization of growth, as well as dice-coefficient between manual and machine-derived, longitudinally.