Myelin mapping is a method for "myelin" quantification (Glass et al., however see Uddin et al., who suggest it does not represent myelin specifically, but can still be used to examine white matter microstructure [aka you can still use the method to tell if there's something going down in the white matter]).
The method is simple:
- Get t1w & t2w images in the same space.
- t1w / t2w
- Myelin map
The pipeline expands on the simple method by using the method outlined by Ganzetti et al. (with a couple of tweaks, like using ANTs):
- Warping MNI template to subj space
- Warping MNI eye + temporal bone + brain masks from MNI to subj space
- T2 registered to T1 image through rigid registration
- N4 bias correction on T1 and T2 images
- Linear intensity adjustment of subj space t1 & t2 images w/ eye + bone masks.
- Creation of myelin map (co-registered, bias corrected, intensity adjusted t1w image / co-registered, bias corrected, intensity adjusted t2w image)
- Warping of myelin map from subj space to MNI.
- Smoothing of output.
This repo contains:
-
A notebook showing the raw work / validation of masks.
-
Scripts to create the myelin maps. Functions in the script use Nibabel, Advanced Normalization Tools (ANTs), scipy and some numpy tinkering.
-
MNI templates for t1 and t2 images (in the resources directory)
-
Eye and temporalus muscle masks in MNI space (in the resources directory).
Dependancies:
- Python 3
- Numpy >= 1.14.6
- Scipy >= 1.0.0
- Nibabel >= 2.3.0
- Seaborn >=0.8.1
- Nipype >= 1.1.1
- ANTs >= 2.2.0
Refs:
Ganzetti et al. (2014). Whole brain myelin mapping using T1- and T2-weighted MR imaging data
Glass & Van Essen (2011). Mapping human cortical areas in vivo based on myelin content as revealed by T1- and T2-weighted MRI
Uddin et al (2017). Can T1w T2w ratio be used as a myelin‐specific measure in subcortical structures? Comparisons between FSE‐based T1wT2w ratios, GRASE‐based T1w T2w ratios and multi‐echo GRASE‐based myelin water fractions