MEG Methods Clinics

Once per month, CIBR organizes training in data analysis to MEG/EEG researchers. Each Method Clinic session focuses on a single problem / method / software, with the aim of facilitating usage of new data analysis tools. These sessions comprise of a small theoretical instruction, followed by hands-on training in our data analysis class.

More information: Simo Monto

Annual clinic cycle

Please note that the actual clinic cycle may differ from the annual plan below. Please follow the news section in the home page for more up-to-date schedule!

JANUARY: Pre-processing of MEG data

This Clinic will focus on the first pre-processing steps of the raw MEG data. We will briefly cover the theory behind the main pre-processing steps (mainly Maxfiltering/SSS and ICA) and then go to hands-on practice, testing a few different approaches with real data. I will also go through a fresh, ready-made Python script, that hopefully can be useful in doing these pre-processing steps in a quick, easy and replicable manner. You can also test the new script yourself and we will try to debug it together.

FEBRUARY: Using and importing MR images

In this training, we will learn how (and why) should we use individual MR images to obtain more accurate anatomical information from MEG data. The steps are described in our Intranet instruction pages, so this training will focus on demonstration and hands-on. We will start from how to read the optical MRI disks from Synlab and end with 3D visualization of the final head model.

MARCH: Elekta/MEGIN data analysis suite

On this MEG Methods Clinic, we will focus on the data analysis software provided by Megin, our MEG device manufacturer (previously Neuromag and Elekta). We will mostly practice (single-)dipole fitting with the Source modeling module ("XFit") but also check Data plotting ("XPlotter") and MEG/MRI integration ("MRILab"). MaxFilter pre-processing will not be covered, because it has been dealt with in a separate pre-processing training (also available on the CIBR servers). The Megin DANA suite provides an easy access to checking your evoked responses, doing source modelling and visualising the modelled dipole sources on MRI slices.

APRIL: Meggie GUI for MEG processing

Meggie is a GUI to MNE-Python that is produced and maintained in-house. With Meggie, you can easily do the most common MEG data analyses, including pre-processing, averaging, frequency and TF analysis, and cross-subject analysis.

MAY: Basics of MNE-Python

For maximal flexibility, reproducibility and efficiency, researchers most often learn to program simple scripts for their data analysis. In this introductory session, we will cover the basics of MNE-Python. We will start by reviewing the possible Python environments in our servers, and learn to write and execute an elementary analysis including data reading, filtering, epoching, averaging and visualisation. From here it is easy to extend analysis with the help of the software documentation and example scripts.

SEPTEMBER: MEG source modeling

We will study how to perform linear distributed source modeling (MNE, sLORETA) with the MNE software (both GUI and scripting). A review of the basics of source modeling is included in the session. Note that dipole source modeling is included in the Megin DANA training.

OCTOBER: Beamforming of evoked and ongoing data

Together with Jan Kujala. We will study how to apply beamforming methods to evoked responses as well as in the frequency domain for event-related and on-going data.

NOVEMBER: Advanced analysis methods I

Description will follow.

DECEMBER: Advanced analysis methods II

This training gives a theoretical introduction and practical hands-on training to multivariate classification methods, i.e. MEG decoding. We will utilize high-level scripting (= brief, simple and functional) with the MNE-Python and Scikit-learn software packages. As in all our Method Clinic trainings, the required software and data can be found on the CIBR servers. However, you can have your own MEG/EEG data and laptop if you want. Familiarize yourself with the decoding concept and elementaries of the Python language.