DATA SUMMARY


These data contain MRI (magnetic resonance imaging), behavioural and MEG (magnetoencephalography) data. They are all collected from the same sample of individuals (although a few subjects lack some data because of subject drop-out). The subject numbers have been matched across data types so that the same subject number matches the same person in every dataset. Subjects numbers can be found on the filenames themselves.

The pdf file in the top level directory called "paper_that_uses_these_data.pdf" is Furl N, Lohse M, Pizzorni-Ferrarese F. (2017). Low-frequency oscillations employ a general coding of the spatio-temporal similarity of dynamic faces. Neuroimage 157:486-499. doi: 10.1016/j.neuroimage.2017.06.023

Behavioural data and magnetic resonance imaging data (functional and structural) come in their own zip archives. Each MEG participant/subject is provided in a separate zip archive, most of them >3 GB in size. Although we regret the need to download so many large files, each file is a manageable download by itself. We were concerned one >50G file would take excessive amounts of time and might risk failure to download completely on some people's systems. 


FILE TYPE SUMMARY
	

MRI files are not familiar file types and require a special viewer (an example viewer fMRIcron, is enclosed) to view them in their 3D format. 

MEG files require special processing and specialty knowledge to produce images and so cannot be viewed directly. 

For neither data type is it practical to supply in text format but we have supplied the easiest to read and most commonly used raw data format for these types of data. Please be aware that both file types are very large (especially MEG data). 

Both MRI and MEG experiments are associated with data about the experimental runs. These data are needed for analysing the data (what stimuli were shown at what time, etc). For MRI, text files are supplied. Fopr MEG data, this information is written directly to the MEG data file itself. Stimulus codes can be decoded when converting the data. More information can be found on this process below.


The behavioural data we have supplied in text format.


MEG DATA

MEG data must be read into a specialised data analysis programme. We used SPM http://www.fil.ion.ucl.ac.uk/spm/. The data stores "triggers" which code for the times at which stimuli occur and what stimuli were shown. They are given 36 possible codes. These codes correspond directly to the filenames (e.g., 11, 12, 13, etc.) of the movie stimuli found in ~\behavioural_data\stimuli\from_behaviour_and_meg. 


MRI DATA


~\mri_structural_and_localiser_data\ contains all structural (in MPRAGE folders) and functional (in LOCALISER folders) data. In the paper, structurals were used for purposed of MEG source reconstruction. The functional scans were used to identify face-selective brain areas FFA, OFA, STS, V5, which were used as regions of interest to constrain the MEG source localisations. 

MRI files can be viewed using the enclosed viewer MRIcron (we encloded the most recent Windows-compatible version). Instructions for installing and using MRIcron can be found at http://people.cas.sc.edu/rorden/mricron/install.html. To analyse MRI and MEG, you need more sophisticated softeware. In the paper, we used SPM.http://www.fil.ion.ucl.ac.uk/spm/

Please note that structural scans have been modified slightly to remove data pertaining to the participants' faces, ensuring anonymity (using the SPM de-face algorithm).

~\mri_structural_and_localiser_data\experimental_localiser_data contains eprime files for every fMRI subject describing the events of the localiser run. These data are summarised in the data file summary_data.txt. This summary can be used for fMRI data analysis.


BEHAVIOURAL DATA


\behavioural data contains data from the same subjects, who made form similarity judgements and motion similarity judgements for pairs of the 36 movies (see Stimuli folder for the movies). These are the same movies as shown to participants during MEG scanning and the paper compares the behavioural similarity with the MEG response similarity between pairs of movies.

The behavioural experimental paradigm was run in a programme called PsychoPy, which can be downloaded for free at http://psychopy.org/installation.html. The PsychoPy scripts for the paradigm are found in the ~\behavioural_data\psychopy_paradigm_programme directory. The programmes is not guranteed to work with versions of PsychoPy later than v1.85.2.

Inside ~\behavioural_data\data, we have enclosed all the files created during behavioural experiment runs for individual participants, including the psychopy output and comma and space delimited text files. We also include a tab-delimited textfile that combines all the individual subject data into one file called behavioural_data.txt.