![Oxford-MM-1 T1, T2 FLAIR, and DTI volumes][1]*Oxford-MM-1 T1, T2 FLAIR, and DTI volumes*
![Oxford-MM-1 T2 FLAIR overlayed with DTI volume][2]*Oxford-MM-1 T2 FLAIR overlayed with DTI volume*
Online viewing (Firefox, or Google Chrome recommended): https://open.win.ox.ac.uk/pages/fsl/oxford-mm-templates/
Version 1 of our multimodal asymmetric template with its T1, T2 FLAIR and DTI volumes. It was constructed from **240 UK Biobank individuals uniformly sampled from the age range 50-55 years (50% female/male)**. We believe a template from this age range will be useful for a wealth of studies.
Note that the template was **rigidly transformed to MNI space** to capture the population’s average brain shape and size, and avoid the introduction of unrepresentative scaling effects.
When using this template please cite:
**Arthofer, C., Smith, S.M., Jenkinson, M., Andersson, J., Lange, F. Multimodal MRI template construction from UK Biobank: Oxford-MM-0. (2021). Organisation for Human Brain Mapping (OHBM)**
Several improvements to the pipeline include the use of:
- weighted brain masks in T1 and DTI reference space, as well as in subjects' native DTI spaces to avoid the introduction of noise around
- the brain FSL's epi_reg with BBR for more accurate rigid registrations between dMRI and T1 median (instead of mean) template
- for the first few iterations to provide a better initialization non-defaced images
Our general pipeline for template construction can be downloaded from https://git.fmrib.ox.ac.uk/cart/mm-template-construction.
The images and transformations below are available as part of this package. Image dimensions are 182 x 218 x 182, voxel size is 1 mm isotropic and bit depth is 32 bit.
| File | Description |
|---|---|
| `OMM-1_T1_head.nii.gz` | T1 whole-head volume
| `OMM-1_T1_brain.nii.gz` | Brain-masked T1 volume
| `OMM-1_T2_FLAIR_head.nii.gz` | T2 FLAIR whole-head volume
| `OMM-1_T2_FLAIR_brain.nii.gz` | Brain-masked T2 FLAIR volume
| `OMM-1_DTI_tensor.nii.gz` | 4D tensor volume
| `OMM-1_DTI_FA.nii.gz` | Fractional anisotropy
| `OMM-1_DTI_MD.nii.gz` | Mean diffusivity
| `OMM-1_DTI_MO.nii.gz` | Mode of the anisotropy
| `OMM-1_DTI_L1.nii.gz` | 1st eigenvalue
| `OMM-1_DTI_L2.nii.gz` | 2nd eigenvalue
| `OMM-1_DTI_L3.nii.gz` | 3rd eigenvalue
| `OMM-1_DTI_V1.nii.gz` | 1st eigenvector
| `OMM-1_DTI_V2.nii.gz` | 2nd eigenvector
| `OMM-1_DTI_V3.nii.gz` | 3rd eigenvector
| `OMM-1_T1_brain_mask_average.nii.gz` | Average (soft) brain mask by averaging over all resampled individual masks
| `OMM-1_T1_brain_mask_weighted.nii.gz` | Weighted (soft) brain mask for registrations with whole-head images only; reduces regularisation in the brain compared to skull and surrounding regions (can be used as --mask_ref_scalar input to MMORF)
| `OMM-1_DTI_mask_average.nii.gz` | Average brain mask for the DTI volume
| `OMM-1_DTI_mask_weighted.nii.gz` | Weighted brain mask for the DTI volume to reduce the impact of noise around the edges of the brain during registrations (can be used as --mask_ref_tensor input to MMORF)
| `transformations/FSL-MNI-152_to_oxford-mm-1_warp.nii.gz` | Combined warp (affine + nonlinear) to transform images (e.g., masks) from the space of FSL’s MNI 152 template to the space of the OMM-1 template. Directly compatible with FSL’s applywarp and vecreg function.
| `transformations/oxford-mm-1_to_FSL-MNI-152_warp.nii.gz` | Combined warp (affine + nonlinear) to transform images (e.g., masks) from the space of the OMM-1 template to the space of FSL’s MNI 152 template. Directly compatible with FSL’s applywarp and vecreg function.
DOI 10.17605/OSF.IO/S9GE4
Copyright, 2022, University of Oxford. All rights reserved
[1]: https://git.fmrib.ox.ac.uk/fsl/oxford-mm-templates/-/raw/main/Oxford-MM-1/oxford-mm-1_overview.jpg
[2]: https://git.fmrib.ox.ac.uk/fsl/oxford-mm-templates/-/raw/main/Oxford-MM-1/oxford-mm-1_T2+tensor.png
