For a more up-to-date repo on the same subject please look at AD-DL repo.

Before launching this code or one of your own you should create a conda env

You can install miniconda on Linux with the following commands:

$ curl https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh -o /tmp/miniconda-installer.sh
$ bash /tmp/miniconda-installer.sh

Type yes when asking to add the miniconda path to your path and restart your session

You can now create your environment and install all the recquirements with:

$ conda create -n deep_ADNI python=3.6
$ git clone https://github.com/14thibea/deep_learning_ADNI.git
$ pip install -r deep_learning_ADNI/recquirements.txt

You also need to install pytorch. Please see Pytorch installation in order to choose the correct command.

You can train a network by typing:

$ python main/network.py tsv_path results_path caps_path