# Final warp computed is not diffeomorphic

**URL:** https://community.mrtrix.org/t/final-warp-computed-is-not-diffeomorphic/1658
**Category:** Uncategorized
**Created:** [May 24, 2018, 11:15am UTC](https://community.mrtrix.org/t/final-warp-computed-is-not-diffeomorphic/1658 "2018-05-24T11:15:40Z")
**Posts on this page:** 1
**Showing post:** 3

<div class="post-metadata">

### Author: ![maxpietsch](https://community.mrtrix.org/user_avatar/community.mrtrix.org/maxpietsch/32/25_2.png) [@maxpietsch](https://community.mrtrix.org/u/maxpietsch)
#### Post date: [October 3, 2019, 10:25am UTC](https://community.mrtrix.org/t/final-warp-computed-is-not-diffeomorphic/1658/3 "2019-10-03T10:25:37Z")

</div>

If the images are not ODFs, I’d recommend removing large bias fields if present, rescaling the images to roughly 0.5 average intensity but most importantly match intensities across input images. [This](https://community.mrtrix.org/t/poor-registration-with-mrregister-compared-with-ants/1233/7) might help.

You can also use a different cost function for instance in ANTs and convert the warps to mrtrix format.

Make sure the affine registration and masks are sensible.

You can convert the non-diffeomorphic warp to the jacobian determinant image and check where the Jacobian is far from 1. You could exclude these areas from the mask, forcing registration to ignore those.

---

_[View the full topic](https://community.mrtrix.org/t/final-warp-computed-is-not-diffeomorphic/1658)._