Skip to content

Project your functional brain signal onto the white matter, and explore the pathways supporting brain functions.


Notifications You must be signed in to change notification settings


Repository files navigation


The Functionnectome is a python package which apply the eponym method and combines the functional signal from distant voxels of the classical fMRI 4D volume using their probabilistic structural relationship given by anatomical priors of the involved brain circuits.

To launch the Functionnectome, if it was properly installed with the pip command (see Installation below), open a terminal and run the command FunctionnectomeGUI to use the GUI, or Functionnectome <setting_file.fcntm> if you already have a settings file ready.


WhiteRest is a tool to explore the potential impact of white matter lesions on resting-state networks. It is installed along the Functionnectome and is based on the WhiteRest atlas. More information is available in the WhiteRest manual.


The complete manual is available of the project's GitHub page ( under the name The_Functionnectome_Userguide.pdf

System Requirements


The Functionnectome can run on a standard computer depending on the type of analysis. A voxel-wise analysis would require at least 4 (or better, 12 or 16) CPUs to shorten the run-time. A region-wise analysis should run on almost any computer.

Operating system

The Functionnectome is supported for MacOS and Linux. It might work on Windows but has yet to be tested. The package has been tested on Linux (Centos7) and MacOS. Test on Windows will come later.

Python Dependencies

The package mostly requires default Python libraries. It also depends on classic scientific libraries:



Using command lines:

  • pip install git+


  • python -m pip install git+

It is also possible to manually download the scripts from the GitHub and run them. More details are available in the manual.


Using command lines:

  • pip install --upgrade git+


  • python -m pip install --upgrade git+

Run time

The run time may vary a lot depending on many different factors (from hardware to input data). As a rule of thumb, the voxel-wise analysis can run on a dataset in under 10 to 20 min on a typical desktop computer. The region-wise analysis used to be faster but I'm not sure it is still the case (haven't tested in a while).

Using the package

After installing the package, enter the FunctionnectomeGUI command in the terminal to start using the interface.

A quick guide and a comprehensive guide with step by step are documented in the manual.

Using the scripts directly

The main program,, runs the computation. It requires one argument: the path to a text file (with the ".fcntm" extension) containing the properly formated input parameters. The .fcntm file can be easily created using the GUI accompanying the package: