TNM_project

Building the models:

1. Install the necessary python packages by running

    pip install -r requirements.txt
   

2. (Optional) Install the matlab engine by following the steps outlined here OR execute the matlab script manually when prompted (not recommended and very tedious).

3. Download the datasets ds003059 and ds00030 to the data/ directory.

4. (Optional) change the parcellation scheme by modifying the corresponding entry in config.yml. Options include:

   • yeo for Yeo-2011
   • harvox for Harvard-Oxford
   • aal for Automated Anatomical Labeling
   • schaefer for Schafer-2018

5. Execute main.py.

Reproducing the data analysis:

1. Make certain the necessary files are present: Either build the models yourself (follow the steps above) or download our pre-built models by running

    curl https://polybox.ethz.ch/index.php/s/jioOccqPFjU6cpk/download -o output_DCM.7z
   

   and unpacking to a folder called output_DCM/.

2. Run analysis.m using MATLAB.