Which datasets are used in the documentation and where are they used?

The FieldTrip website and ftp server include a variety of datasets. Some of these are used in the tutorials, while others are used in example MATLAB scripts. All of these datasets were recorded from humans, unless otherwise specified.

Below you will find a brief description of each dataset. The name given to each dataset is meant to provide an instant understanding of 'how the data was collected', 'the cognitive domain to which the data relates', and if it applies, 'what specific topic within the domain'. For example: 'MEG-language-semantics' will tell you that the data was collected in an MEG scanner, relates to research in the language domain, and specifically to semantics. Links to the FieldTrip ftp server are provided so that you can download the dataset of interest.

  • Dataset specifically generated by FieldTrip members for teaching artifact rejection.
  • Data composed of 10s segments with no gaps between segments.
  • Download the datasetArtifactMEG.zip from our FTP server.
  • Tutorials using this dataset:
    2010/09/15 15:08 Cristiano Micheli
  • Both wrists in isometric extensions to exert a constant force against a lever. Indicate detected change (speed) of visual stimulus by further extending cued hand. Analyzed the gamma-band coherence as an indicator of communication between sensorimotor regions and and motorneurons in the forearm.
  • 275 CTF MEG system
  • Download the publication by Schoffelen et al.(2011).
  • Download the dataset SubjectCMC.zip from our FTP server.
  • Tutorials using this dataset:
  • Covertly attend to cued square: left or right (hemifield). Used alpha oscillations in covert attention as a classifier (left vs. right) for use in brain computer interface.
  • 275 CTF MEG system
  • Download the dataset covatt.mat from our FTP server.
  • Link to the publication by van Gerven et al.(2009).
  • Tutorials using this dataset:
  • Dutch words presented in visual form, auditory form, and picture form. Multivariate pattern analysis used to decode conceptual representations from event-related data
  • 64 channel ActiCap system (Brain Products GmbH)
  • Link to publication by Simanova et al.(2010).
  • Download the dataset SubjectEEG.zip from our FTP server
  • Tutorials using this dataset:
  • Dutch words presented in visual form, subjects made positive/negative or animal/human judgments on the nouns.
  • 64 channel ActiCap system (Brain Products GmbH)
  • Download the dataset here from our FTP server
  • Tutorials using this dataset:
    2015/11/11 17:21 Robert Oostenveld
  • Data specifically generated by FieldTrip team members for teaching TMS/EEG data analysis.
  • Transcranial magnetic stimulation applied to primary motor cortex (M1) while subject contracts or relaxes contralateral hand.
  • C-B60 butterfly coil connected to a MagPro X100 (Magventure) stimulator & 61 channel TMS-compatible EEG cap (EasyCap)
  • Download the dataset sp_motor.zip from our FTP server.
  • Tutorials using this dataset:
    2013/09/11 16:40  
  • EEG, MEG, fMRI and structural MRI data from 16 subjects, also used in SPM8 and SPM12 documentation
  • Recorded by Rik Hanson and colleagues in Cambridge, using 306-channel Elekta/Neuromag system
  • Subjects watch familiar, unfamiliar and scrambled faces
  • Download the dataset from the MRC-CBU FTP server.
  • Details mmfaces
  • Tutorials using this dataset:
    2014/12/23 10:19 Robert Oostenveld
    2014/12/23 10:20 Robert Oostenveld
    2014/12/23 10:20 Robert Oostenveld
    2014/12/18 10:48 Robert Oostenveld
  • ECoG data recorded at the Comprehensive Epilepsy Center of the New York University School of Medicine and processed in collaboration with members of the Multisensory Integration Research Group (Charité - University Medicine Berlin).
  • The data was recorded from an epilepsy patient implanted with a subdural grid for pre-surgical evaluation and includes a visual localizer task. Different types of visual stimuli were presented.
  • Download the dataset SubjectNY394.zip
  • Intracranial EEG data (ECoG and Stereo EEG) obtained at the UC Irvine Medical Center. Includes a pre-implantation MR and a post-implantation CT, the latter showing the recording electrodes.
  • Download the dataset SubjectUCI29.zip
  • Tutorials using this dataset: Analysis of human ECoG and sEEG recordings

The Human Connectome Project contains MEG data from close to 100 subjects, which were scanned on the 248-channel 4D/BTi MEG system at SLU, St Louis. The MEG data includes resting-state and three different experimental tasks, and anatomical, functional (1200 subjects, resting and task) and associated behavioral and demographic data is available. See also doi: 10.1016/j.neuroimage.2013.05.056 and doi: 10.1016/j.neuroimage.2015.04.046.

The Cambridge Centre for Ageing and Neuroscience released raw data from the Cam-CAN cohort, including MEG, MRI and cognitive data from approximately 650 males and females uniformly distributed from 18 to 88 years of age. The MEG data (in FIF format) include 3 runs of 1) rest, 2) a sensorimotor task, and 3) a passive sensory task (with separate visual and auditory stimuli).

The Open MEG Archive (OMEGA) is a collaborative effort by the McConnell Brain Imaging Centre and the Université de Montréal. It consists of a open repository with MEG data. This continuously expanding repository also contains anatomical MRI volumes, demographic and questionnaire information, and eventually will feature other forms of electrophysiological data (e.g. EEG, field, and cell recordings). See also doi: 10.1016/j.neuroimage.2015.04.028

The BNCI Horizon 2020 consortium hosts a repository of datasets from brain-computer interface (BCI) and decoding experiments available for free download. The datasets include EEG, fNIRS, and ECoG data collected mainly by the consortium partners in several European countries.

The Kymata Atlas also shares datasets with electromagnetic measurements of the human cortex done at the MRC Cognition and Brain Sciences Unit of Cambridge University. The shared data comprises (averaged) EEG and MEG sensor data, and current density reconstructions. The participants are healthy human adults listening to the radio and/or watching films.

Wakeman and Henson have shared A multi-subject, multi-modal human neuroimaging dataset that includes raw MEG and EEG data from 16 subjects. This data also forms the basis for the SPM documentation.

The Brain Signals website has a list of references to shared EEG data sets.

Other MEG data analysis toolboxes (MNE, BrainStorm) also share tutorial datasets.

Kahana and Rizutto have released EGoG data from 149 subjects collected during Phase I of the Restoring Active Memory (RAM) project. The dataset includes information from 700 sessions, and for every patient, intracranial recording files from 100 to 200 electrode channels, neuro-anatomical information indicating the location of each electrode, precise records of patient behavior and the experimental design documents.

A collection of intracranial animal datasets is shared on the Collaborative Research in Computational Neuroscience data sharing website.