Select Page

BFP: BrainSuite Functional Pipeline

The BrainSuite fMRI pipeline (BFP) is an open-source software workflow for processing raw resting fMRI data. The pipeline processes resting fMRI and anatomical (T1) imaging data using a combination of software that includes BrainSuite, AFNI, and FSL, as well as MATLAB scripts.
The Github Development page for BFP can be found at https://github.com/ajoshiusc/bfp.
BFP is now available as part of the BrainSuite BIDS App.

To facilitate interaction across software packages, the processed fMRI data are represented in a common grayordinate system.

Unique features of the BFP pipeline include cortically-constrained volumetric registration, global PDF-based non-local means filtering (GPDF), and BrainSync, a method for temporal synchronization of resting fMRI data across subjects.

By representing data with respect to the BCI-DNI (combined surface/volume atlas) and the HCP’s grayordinate representation, we aim to facilitate the use of BFP with existing analysis tools.

Features of BFP
Some of the unique features of BFP are shown below:

System Requirements

The current version of BFP is Linux only. A Debian based Linux distribution (Ubuntu/Mint/Debian), 8GB RAM recommended. For Windows users, you can try using a virtual machine, but we have not done testing on such configurations, so we do not support it at the moment.

BFP is now available as part of the BrainSuite BIDS App.

Installation and Setup

  • Install AFNI (Ver. Jun 12 2017 or newer) and FSL 5.0. We recommend using NeuroDebian, as it makes the installation process easier. BFP has been tested with this configuration, but it should work for other versions of Linux and other softwares. However we recommend and support the above mentioned versions.

  • Install BrainSuite 18a.

  • Install Matlab 2017a if you want to run it from Matlab
  • Download bfp, unzip
  • Open bfp.mtx in Matlab Live Notebook and edit User Inputs section.
  • open <bfpdir>/supp_data/config.ini and set paths for FSL, AFNI and BrainSuite.
  • For preprocessing fMRI, run src/preproc/bfp.mtx

Usage

Matlab

bfp(configfile,t1,fmri,studydir,subid,sessionid,TR)

configfile: Configuration file that you edited during installation, (<bfpdir>/supp_data/config.ini)

t1: T1 image in NIFTI-1 format e.g. ‘sub01-bT1w.nii.gz’

fmri: string or cell array of fMRI NIFTI-1 formatted files e.g. ‘sub_01_rest_bold.nii.gz’

studydir, subid: Outputs will be saved in the studydir/subid for each subject e.g. studydir=’mystudy’, subid=’sub-01′

sessionid: string or cell array of session id for the input fmris e.g. ‘rest’ for resting state

e.g., bfp /home/ajoshi/bfp/supp_data/config.ini /home/ajoshi/sub-01_T1w.nii.gz /home/ajoshi/sub-01_rest.nii.gz /home/ajoshi/mystudy sub-01 rest 2

Binary

You can run BFP from the bash shell using the command mentioned below.

bfp.sh /home/ajoshi/bfp/supp_data/config.ini /home/ajoshi/sub-01_T1w.nii.gz /home/ajoshi/sub-01_rest.nii.gz /home/ajoshi/mystudy sub-01 rest 2