wxyabc/nbest repository overview

This docker repository contains a set of scripts for automated preprocess of non-human primates T1w MRI.

nBEST: non-human Primates Brain Extraction and Segmentation Toolbox: A convenient, precise, and fully automated AI pipeline for preprocessing NHPs from multi-species, multi-site, and multi-developmental stages. It can realize the entire preprocess for raw T1w images with one instruction without any manual intervention.

The Nvidia-docker is required since our pipeline needs GPU support. Please refer to the official installation guideline to install Nvidia-docker.

Pull the docker image:

docker pull wxyabc/nbest:1.3

The container is totally free. Please first contact [email protected]⁠ to get a free license. Then create a folder (License), and put the License.txt into this folder. Some examples from multi species NHPs have been provided in demo/data. Run the demo and see results in demo/:

docker run -it --gpus=all --ipc=host wxyabc/nbest:1.3

To preprocess your own T1w raw data, create a directory named 'Macaque_testing_scans' (data_folder), which will be mounted to the container. The T1w scans that need to be processed should be in 'Macaque_testing_scans/data/xxxxx.nii.gz' Run by a new terminal:

docker run -it --gpus=all --ipc=host -v /absolute/path/to/Macaque_testing_scans:/workspace/demo/ wxyabc/nbest:1.3 cp nBEST.py demo/nBEST.py cd demo

If you want to process your data with all three steps (i.e., brain extraction, cerebrum&brainstem removal, and tissue segmentation), run:

python nBEST.py

And if you want to process your data without brain extraction (which means your input should be brain images without skull), run：

python nBEST.py -skip_be

--Output illustration--

Upon completion of the processing, all resulting outputs will be generated and stored in the Macaque_testing_scans directory. The following explains what the results are:

brain_img/: The skull stripped brain image. brain_mask/: The brain mask of raw image. brain_cerebrum/: The cerebrum (remove cerebellum and brainstem from brain image) image. brain_cerebrum_mask/: The cerebrum mask of raw image. brain_tissue/: The cerebrum tissue segmentation. brain_subcortical/: The subcortical segmentation.

We welcome the sharing of any brain MRI data from NHPs to continuously update our toolbox. We will strictly adhere to the relevant data usage protocols and provide data contributors with both processed results and manual corrections, leveraging the expertise of our seasoned professionals. We look forward to connecting!