Mr-90-Style/biananes

Scalable fMRI Data Analysis, <Big Data Approaches for the Analysis of Large-Scale fMRI Data Using Apache Spark and GPU Processing: A Demonstration on Resting-State fMRI Data from the Human Connectome Project> thsis source code

biananes

Scalable fMRI Data Analysis

biananes is a library for fMRI data analysis on large datasets running on the Apache Spark framework.

Installation

A) Apache Spark: For the installation and setup of a Spark environment see the Apache Spark documentation.

B) Spark NiftiReader C library

All nodes running Spark workers need to have the libsparkniftireader installed

    # dpkg -i libsparkniftireader-0.0.2.deb

Note: libsparkniftireader depends on libnifti-dev

    # apt-get update
    # apt-get install libnifti-dev

C) After building the biananes.jar with

    $ mvn package

Usage

One could add the JAR to a spark-shell

    ./bin/spark-shell --master local[4] --jars /path/to/jar/biananes.jar

and start using it:

    scala> import org.biananes.io.NiftiTools.NiftiImageReadMasked
    scala> val hcp_root = sys.env("HCP_ROOT")
    scala> val img_file = hcp_root + "167743" + "/MNINonLinear/Results/rfMRI_REST1_RL/rfMRI_REST1_RL.nii.gz"
    scala> val mask_file = hcp_root + "167743" + "/MNINonLinear/Results/rfMRI_REST1_RL/brainmask_fs.2.nii.gz"
    scala> val mat = NiftiImageReadMasked(img_file, mask_file, sc)