Introduction

BRAPH is an object-oriented toolbox written in MatLab that uses graph theory to characterize brain connectivity. BRAPH permits one to calculate brain connectivity matrices from various kinds of neuroimaging techniques, including structural magnetic resonance imaging (MRI), functional magnetic resonance imaging (fMRI), electroencephalography (EEG), and positron emission tomography (PET). Once these networks have been built, BRAPH can calculate several graph theory measures.

BRAPH provides a graphical user interface (GUI) that guides the user through all the steps of an analysis of brain connectivity using graph theory: definition of a brain atlas; construction of the connectivity matrices; calculation of global and local measures; comparison between groups and with random graphs using permutation tests; and, finally, visualization of the results. The initial GUI is presented in figure 1.

Figure 1: Initial GUI that appears when BRAPH is launched. From this GUI, it is possible to select the neuroimaging modality (checkboxes on the bottom right) and to launch the software to deal with the corresponding brain atlas, cohort or graph analysis (buttons on the right).
Figure 1: Initial GUI that appears when BRAPH is launched. From this GUI, it is possible to select the neuroimaging modality (checkboxes on the bottom right) and to launch the software to deal with the corresponding brain atlas, cohort or graph analysis (buttons on the right).

Initial GUI that appears when BRAPH is launched. From this GUI, it is possible to select the neuroimaging modality (checkboxes on the bottom right) and to launch the software from different stages of the workflow: brain atlas, cohort or graph analysis (buttons on the right).

In this chapter, we briefly introduce the overall architecture of BRAPH.

While the current version of BRAPH does not allow analyzing data from diffusion tensor imaging (DTI), this will be available in future releases.

Getting Started
  1. Download BRAPH from http://braph.org/software/.
  2. Unzip the downloaded file into the desired directory.
  3. Launch MatLab and change the current folder to the directory chosen in step 2.
  4. Execute braph.m by typing braph in the command line panel of MatLab. This loads all the files necessary to use BRAPH and opens the initial GUI shown in figure 1.

 

Initial GUI

The initial GUI Braph is shown in figure 1 and consists of three main work areas:

  1. Animation panel (on the left). By default, it shows an animation of a sample graph on a brain surface. Alternatively, by checking the corresponding checkbox in the bottom left, it can also feature a slide show of some select graph measures.
  2. Imaging modality selection (at the bottom). A set of four checkboxes permits the user to choose the imaging modality corresponding to the data to be analyzed. Currently, BRAPH can analyze:
    • structural magnetic resonance imaging (MRI) data;
    • functional magnetic resonance imaging (fMRI) data;
    • positron emission tomography (PET) data;
    • electroencephalography (EEG) data.
  3. Workflow panel (on the right). Series of push buttons that permit to launch various stages of the graph analysis for the selected imaging modality. For example, for MRI:
    • Brain Atlas defines the atlas used in the analysis;
    • MRI Cohort imports the MRI data of the subjects;[1]
    • MRI Graph Analysis performs the brain connectivity analysis.[2]

General workflow

The workflow of BRAPH is organized along a series of GUIs that guide the user into the analysis of brain connectivity. This is illustrated in figure 2.

Figure 2: General workflow of BRAPH for MRI, fMRI, EEG, and PET data analysis. Each box represents a GUI available in BRAPH, while the arrows indicate the typical workflow. The formats of the BRAPH files are indicated within each box. The abbreviation WU refers to weighted udirected graphs, BUD to binary undirected graphs at a fixed density of connections, and BUT to binary undirected graphs at a fixed threshold.
Figure 2: General workflow of BRAPH for MRI, fMRI, EEG, and PET data analysis. Each box represents a GUI available in BRAPH, while the arrows indicate the typical workflow. The formats of the BRAPH files are indicated within each box. The abbreviation WU refers to weighted undirected graphs, BUD to binary undirected graphs at a fixed density of connections, and BUT to binary undirected graphs at a fixed threshold.
  1. Brain Atlas permits the user to create and manage a brain atlas.
  2. MRI/fMRI/PET/EEG Cohort permits the user to create a cohort of subjects and upload their data.
  3. MRI/fMRI/PET/EEG Graph Analysis permits the user to define the parameters for the graph analysis. After deciding the type of analysis, the user is redirected to a specialized GUI:
    • MRI/fMRI/PET/EEG Graph Analysis WU to perform a graph analysis using weighted undirected graphs.
    • MRI/fMRI/PET/EEG Graph Analysis BUD to perform a graph analysis using binary undirected graphs at a fixed density of connections.
    • MRI/fMRI/PET/EEG Graph Analysis BUT to perform a graph analysis using binary undirected graphs at a fixed threshold.

 

Example data

MRI and fMRI data sets are provided on the website http://braph.org. They are provided only for training purposes. They have been randomly generated and do not correspond to real clinical data.

Footnotes and references

  1. ^ Depending on the selected imaging modality, this can also be fMRI Cohort, PET Cohort, or EEG Cohort.
  2. ^ Depending on the selected imaging modality, this can also be fMRI Graph Analysis, PET Graph Analysis, or EEG Graph Analysis.