Resources:

Human Connectome Project: http://www.humanconnectomeproject.org/

MEG pipeline: https://www.humanconnectome.org/software/hcp-meg-pipelines

Data download : Only needed if you cannot find the data locally. https://db.humanconnectome.org (Does not work with Firefox (April 2018)) We recommend getting in touch with Maximilien CHAUMON instead.

FAQ: https://wiki.humanconnectome.org/display/PublicData/HCP+Users+FAQ

Prerequisite

We hold a local copy of a good portion of the MEG data of the HCP. This data is located on the network at /this/address/ in original zip (archive) format. If the data you are looking for is not at this address, please contact Maximilien CHAUMON.

Goal of this tutorial

Create simple code to extract parts of the HCP data to a different folder.

Step-by-step guide

All you need to do to extract the data is to:

Locate the zip files, open one of them and look for the data you need
List the zip files (e.g. for all subjects)
Unzip what you need locally.

Locate the files you need. These are named, according to the HCP nomenclature. For instance, the resting state data is named like this:

*_MEG_Restin_preproc.zip

The folder structure can be explored by opening one zip file. Folder names are rather explicit (see right).

For instance, here, I want to retrieve all the resting state preprocessed MEG files. I can see that all of them contain the pattern *Restin_rmegpreproc.mat

List Zips and extract

rootdir = '/where/the/data/is';
destinationdir = '/where/the/data/goes';

% the patterns below should be adapted to capture specific zip files and specific files within it
zippattern = '*_MEG_Restin_preproc.zip';
filepattern = '*-Restin_rmegpreproc.mat';

cd(rootdir)
zips = dir(zippattern);
cd(destinationdir)
for i = 1:numel(zips)
	zipfname = fullfile(zips(i).folder,zips(i).name);
	system(['unzip -n ' zipfname ' ' filepattern]);
end

In the zip file:

After unzip in your local folder:

Related articles

Page:
Mapping HCP MEG sources to an atlas (CENIR)
Page:
Retrieve Human Connectome Project MEG data (CENIR)

Browser not supported