DL+DiReCT combines a deep learning-based neuroanatomy segmentation and cortex parcellation with a diffeomorphic registration technique to measure cortical thickness from T1w MRI.
If you are using DL+DiReCT in your research, please cite (bibtex) the corresponding publication:
Rebsamen, M, Rummel, C, Reyes, M, Wiest, R, McKinley, R.
Direct cortical thickness estimation using deep learning‐based anatomy segmentation and cortex parcellation.
Human brain mapping. 2020; 41: 4804-4814. https://doi.org/10.1002/hbm.25159
Download and install Miniconda and create a new conda environment:
conda create -y -n DL_DiReCT python=3.10
source activate DL_DiReCT
cd ${HOME}
git clone https://github.com/SCAN-NRAD/DL-DiReCT.git
cd DL-DiReCT
pip install numpy && pip install -e .
Run dl+direct on a T1-weighted MRI including skull-stripping (--bet) using HD-BET with:
source activate DL_DiReCT
dl+direct --subject <your_subj_id> --bet <path_to_t1_input.nii.gz> <output_dir>
Following files of interest are generated in the output directory:
- T1w_norm.nii.gz Re-sampled input volume
- T1w_norm_seg.nii.gz Segmentation
- T1w_norm_thickmap.nii.gz Thickness map
- result-vol.csv Segmentation volumes
- result-thick.csv ROI-wise mean cortical thickness
- result-thickstd.csv ROI-wise standard deviations of cortical thickness
- label_def.csv Label definitions of the segmentation
Results may be collected into FreeSurfer alike statistics files with stats2table
.
To process images with a contrast agent (contrast-enhanced), use the option --model v6
(Rebsamen et al., 2022).
The following models are available with the --model ...
option:
v0
: Default, for un-enhanced T1w MRI, cortex parcellation with Desikan-Killiany atlas (Rebsamen et al., 2020)v6
: For both contrast-enhanced and un-enhanced MRI (Rebsamen et al., 2022)v7
: Same as v6, with 74 region per hemisphere according the Destrieux atlas (Rebsamen et al., 2022)
For further details, consult the corresponding publication and the FAQ or contact us