Free academic toolboxes have gained increasing prominence in MEG analysis as a means to disseminate cutting edge methods, share best practices between different research groups and pool resources for developing essential tools for the MEG community. In the recent years large and vibrant research communities have emerged around several of these toolboxes. Teaching events are regularly held around the world where the basics of each toolbox are explained by its respective developers and experienced power users. There are, however, two knowledge gaps that our proposed workshop aims to address. Firstly, most teaching examples only show analysis of a single ‘typical best’ subject whereas most real MEG studies involve analysis of group data. It is then left it to the researchers in the field to figure out for themselves how to make the transition and obtain significant group results. Secondly, we are not familiar with any examples of fully analyzing the same group dataset with different academic toolboxes to assess the degree of agreement in scientific conclusions and compare strengths and weaknesses of various analysis methods and their independent implementations. The proponents of the present workshop are lead developers of five most popular free academic MEG toolboxes (in alphabetic order): Brainstorm, EEGLAB, FieldTrip, MNE, NUTMEG, and SPM. Ahead of the workshop the research team for each toolbox will analyze the same group MEG/EEG dataset. This dataset containing evoked responses to face stimuli was acquired by Richard Henson and Daniel Wakeman, who won a special award at BIOMAG2010 to make it freely available to the community. All the raw data are available at  

Detailed instructions for each toolbox will be made available online prior to the event including analysis scripts and figures of results. All analyses will show a full pipeline from the raw data to detailed publication quality results.

At the workshop each group will briefly introduce their software and present the key results from their analysis. This will be followed by a panel discussion and questions from the audience.

Following the event we plan to integrate the suggestions and questions from the workshop audience and to publish the analyses details as a special issue in one of the profile journals so that the proposed best practices will be endorsed by peer review and become citable in future publications. Other research groups will be invited to contribute to the special issue as long as they present detailed descriptions of analyses of group data that are freely available online and make it possible for others to fully reproduce their analysis and results.

We hope that this proposal will lead to creation of invaluable resource for the whole MEG community and the workshop will contribute to establishment of good practice and promoting consistent and reproducible analyses approaches. The event will also showcase all the toolboxes and will be of interest to beginners in the field with basic background in MEG who contemplate the most suitable analysis approach and software for their study as well as to experienced researchers who would like to get up to date with the latest methodological developments.

Speakers:

Francois Tadel (McGill Univ., Canada)
“Group analysis in Brainstorm”

Arnaud Delorme (Paul Sabatier Univ., France)
“Group analysis in EEGLAB”

Robert Oostenveld (Donders Inst., The Netherlands)
“Group analysis in Fieldtrip”

Alexandre Gramfort (Telecom ParisTech, France)
“Group analysis in MNE”

Srikanran Nagarajan (UCSF, USA)
“Group analysis in NutMEG”

Richard Henson (Univ. of Cambridge, UK)
“Group analysis in SPM”

"Comparing and combining MEG with EEG data is a very useful approach in analyses of clinical data. MEG often shows better SNR (Signal-to-Noise-Ratio) for superficial tangential sources, however, radial source components are invisible in a spherical head geometry. EEG has a more isotropic sensitivity distribution and is able to measure signals from radial and deep sources. Due to this, SNRs in EEG are often lower as compared to MEG and EEG maps are more difficult to interpret.
Propagating source activity can show phase-shifts between MEG and EEG, the reason also lays in the different sensitivity profiles of MEG and EEG. For example, if an activity starts with a more radial orientation followed by propagation to a location with a more tangential orientation (e.g. from the crown to the wall of a sulcus), the EEG waveforms lead and vice versa (e.g from the tip to the lateral aspect of a temporal lobe).
Comparisons between both modalities from epileptic spike activity and approaches to combine both data by SNR-transformation will be discussed. Pitfalls and considerations about volume conductor properties and extended sources will also be covered by the speakers.

Speakers:

Manfred Fuchs (Compumedics Neuroscan, Germany)
"Source Reconstruction from combined MEG and EEG Data"

Stefan Ramp (Erlangen Univ., Germany)
"Simultaneous and combined EEG in epileptic focus localization: A clinical perspective"

Robert Knowlton (Univ. of California, San Francisco, USA)
"Ictal EEG source localization on the Seizure Monitoring Unit using 31 recording channels"

Michael Wagner and John Ebersole (Compumedics Neuroscan, Germany)
"Benefits of combined MEG/EEG in presurgical evaluation of epilepsy: a study of 250 patients"

In the last decade, MEG data analysis has benefited from the increased availability of computational resources at an affordable price. In particular, source modeling of MEG data has moved from being mostly concerned with the spatial localization of neural sources to the full exploitation of the temporal dimension. Such development is mediated by the use of, e.g., mixed norm regularization methods, Bayesian filtering/smoothing algorithms, and/or sophisticated Monte Carlo sampling techniques.
At the same time, novel experimental approaches with simultaneous recordings of intracranial and surface data allow unprecedented validation of inverse methods in real (not just realistic) scenarios.
We speculate that such advances in the solution of the MEG inverse problem can have a notable impact in two important fields, namely, in the study of source domain connectivity – to which the first part of the symposium is dedicated - and in the study of epilepsy – to which the second part of the symposium is dedicated.
The symposium will summarize recent methodological and experimental advances,  and discuss current and potential impact in the mentioned applications, also trying to point out still un-satisfacted needs that may guide further developments

Speakers:

Erkki Somersalo (Case Western Reserve Univ., USA)
“Bayesian hierarchical models and mapping of deep brain activity in MEG”

Vittorio Pizzella (‘G. d’Annunzio’ Univ., Italy)
“Influence of MEG source estimation on functional connectivity analysis “

Matti Hamalainen (Massachusetts General Hospital, USA / Karolinska Institutet, Sweden)
“Time Matters: Incorporating Temporal Structure into MEG Source Estimation”

Karim Jerbi (Univ. of Montreal, Canada)
“Detection of long-range coupling in MEG source-space: Does the choice of inverse method matter?”

Sara Sommariva (Universita di Genova, Italy)
“Automatic multiple dipole estimation through Semi-Analytic Sequential Monte Carlo samplers”

Christian G. Benar (Aix-Marseille Université, France)
“Simultaneous recordings of MEG and intracerebral recordings for developing and validating signal processing methods”

Cristophe Grova (Concordia Univ. / McGill Univ., Canada)
“Source localization of transient and oscillatory epileptic activity in EEG/MEG using Maximum Entropy on the Mean (MEM)”

Alexandre Gramfort (Universitè Paris-Saclay, France)
“Spatio-temporal imaging with MEG and EEG under sparsity constraints”

Geertjan Huiskamp (Univ. Medical Center Utrecht, The Netherlands)
“Current practice of the use of MEG source localization in epilepsy surgery”

Nicole van Klink (UMC Utrecht, The Netherlands)
“MEG inverse methods to investigate epileptic high frequency oscillations”