Ex 2: MEG evoked responses

This exercises familiarizes you with processing and interpretation of evoked responses.

Download a pre-recorded MEG data set as a Matlab file from here (size is about 220 MB). This file holds three Matlab variables:

  • data : a matrix of MEG-channels (315) by time points (200000)
  • channels : a vector (315) of channel names in the same order as the rows in the data matrix
  • samplingrate : the sampling rate in Hz (the reciprocal of this value is the time between two consecutive time points in the data matrix)

The data are a continuous (non-averaged) recording of a healthy adult volunteer who received 8 kinds of simple sensory stimuli, one at a time, in a random order. Stimulus onsets are marked with trigger events on channel STI014 (you can ignore other STIxxx channels). There is one trigger code corresponding to one stimulus type, and the onset (transition from zero to a non-zero value) of the trigger pulse denotes the onset of the stimulus. The duration of the trigger pulse is irrelevant.

Your task is to:

1. Compute the average evoked response (separately) for the trials with trigger codes 2, 3, and 16 (just ignore the other trigger codes). Choose a time window which includes, say, a 100-ms pre-stimulus baseline and extends to about 500 ms post-stimulus.

2. Visualize the average responses topographically. You may use the plotEF Matlab function available on this page; see Ex 2: Helper functions.

3. Make your best guess which kind of a stimulus corresponds to each trigger code (i.e., which sensory modality and where presented with respect to the body). Exploit the location and timing information; remember that the planar gradiometers peak right above the source. Note that all the stimuli are rather elementary.

4. Write a report. In addition to the text, include three figures, one for each average response. You are encouraged to search relevant literature for typical characteristics of MEG responses to sensory stimuli. Refer to such studies in your report.