nobrainer

Neural networks for brain extraction and labelling from structural magnetic resonance images.

Examples

Please see the examples directory.

Getting started

Get the container

$ docker pull kaczmarj/nobrainer
# or
$ singularity build nobrainer.sqsh docker://kaczmarj/nobrainer

Train your own models

Models can be trained on neuroimaging volumes on the command line or with a Python script. All of the examples can be run within the Nobrainer container. Please see the examples for more information.

Training data pre-requisites:

Volumes must be in a format supported by nibabel.
Feature and label data must be available (e.g., T1 and aparc+aseg).

Training progress can be visualized with TensorBoard:

$ singularity exec --clean-env --bind /path/to/models:/models nobrainer.sqsh \
    tensorboard --logdir /models

Predict using trained models

We are in the process of training robust models for brain extraction and brain labelling. Stay tuned for information on how to use these models.

Funding

The nobrainer project is supported by NIH R01 EB020470.