CiCLONE : Cico Cardinale's Localization Of Neuro-electrodes

Develop

Use poetry (installed via pipx) to setup the virtual env and run it nicely for you.

# Pick a valid Python3 version, e.g. 3.11 or 3.12

$ poetry env use $(pyenv which python3.11)
or
$ poetry env use $(pyenv which python3.12)
then
$ poetry install

Run it

First, ensure that the config.yaml file is present in the config folder in the same directory as main.py and that the output directory exists on your machine. If not copy config.yaml.template to config.yaml and update the fields inside the file.

The pipeline consists of multiple stages that can be run independently or all at once. Each stage contains one or more operations. The available operations are:

• crop - Crop an image file [Input_File, Output_File]
• move - Move a file to a directory [Input_File, Output_Dir]
• coregister - Coregister an image to a reference [Input_File, Reference_File, Output_File]
• subtract - Subtract one image from another [Input_File, Subtract_File, Output_File]
• threshold - Apply thresholding to an image [Input_File, Output_File]
• apply_transformation - Apply a transformation matrix [Input_File, Transformation_File, Reference_File, Output_File]
• extract_brain - Extract brain from an image [Input_File, Output_File]
• mask - Apply a binary mask to an image [Input_File, Mask_File, Output_File]
• register_ct_to_mni - Register CT to MNI space [Input_File, Output_File]
• register_mri_to_mni - Register MRI to MNI space [Input_File, Output_File]
• reconstruct - Run FreeSurfer reconstruction [Input_File, Output_Dir]

The pipeline configuration is defined in YAML format. Each stage has a name and a list of operations. For example:

stages:
  - name: preprocessing
    operations:
      - type: crop
        workdir: preop/ct
        files: ["${name}_CT_Bone", "${name}_CT_Bone_C"]
      - type: move
        workdir: preop/ct
        files: ["${name}_CT_Bone_C", "${subj_dir}/processed_tmp"]

Step 1 : Update the output directory

$ ciclone --update-output-directory /path/to/output/directory

Step 2 : Create a folder for each subject

$ ciclone --subjects subject1 subject<N> --create-folder

Step 3 : Run all stages or a specific one

$ ciclone --subjects subject1 subject<N>

or

$ ciclone --subjects subject1 subject<N> --stages <NAME_OF_THE_STAGE>

Step 4 : Mark your electrodes in 3D Slicer

Here you are on your own. You need to mark your electrodes in 3D Slicer using the r_SUBJECTID_seeg_masked file and your subject CT and/ or MRI and save the coordinates in a JSON file.

Step 5 : Transform coordinates from Subject space to MNI space

$ ciclone --subjects subject1 subject<N> --transform-coordinates /path/to/3D-SlicerJSON/file