Meggie

Meggie instructions

Follow the instruction below to learn the use of Meggie.

You can also visit the Meggie GitHub pages: https://github.com/cibr-jyu/meggie

Initial steps

1. Connect to server via NoMachine or via terminal: ssh -Y username@cibr1.psy.jyu.fi

2. Use command meggie_gui to start Meggie. If it says "command not found", try using commmand "bash" and then meggie_gui again. The main window of Meggie should open.

3. Create a new experiment and then add subjects (actually, raw data files) to this experiment. Subjects in one experiment should be similar to each other (same equipment, same paradigm, etc.). Experiment is created by choosing the File menu and clicking “Create new experiment". Avoid spaces and special characters when naming.

4. Then, add a subject to this experiment. We will use MEG sample data from the MNE software package. Click "Add new..." on the left, locate file /projects/training/MNE/MNE-sample-data/MEG/sample/sample_audvis_raw.fif, and click “Open” and “OK”. Name of the file should soon appear on the list of subjects. The experimental setup is: "In the MEG/EEG experiment, checkerboard patterns were presented into the left and right visual field, interspersed by tones to the left or right ear. The interval between the stimuli was 750 ms. Occasionally, a smiley face was presented at the center of the visual field. The subject was asked to press a key with the right index finger as soon as possible after the appearance of the face."

5. Select the subject from the list and click "Activate selected". Shortly, some info appears on the bottom left of the window, and the analysis tools become available.

6. We can take a look at the data by clicking "Raw plot." You can use arrow keys to move through time or through channels. Note that cardiac artifacts are clearly seen for example on channel MEG 0143. More functionality can be found by clicking the Help button. Close the plot window.

Preprocessing

7. Next, we would set the bad channels by "Customize channels.." on the right. Now bad channels are already set up for us, so we'll skip it this time.

8. Open filter dialog in the section Available actions. Low-pass and high-pass are set to be 1 and 40 as default. Let's go with these and press compute.

[9.-11. Do ICA instead?]

9. Open Calculate ECG projections dialog. This will be used to compute SSP vectors, which can then be used to project away the cardiac artifacts. Select the MEG 0143 channel, where cardiac artifacts were clearly visible, and click "Plot events" to be sure that the artifacts are correctly found. The window that appears show all identified ECG epochs. Here, you can also select the number of SSP vectors computed for each channel type. Then click Compute.

10. When the computation is finished, open Apply ECG projections dialog and select all of the available projections. “Preview” opens a plot that can be used to look how projection vectors affect the data. After plot opens, you can click the "Proj" dialog open from the bottom right of the plot window and choose which projections are applied to visualized data. You need to decide if one SSP vector is effective enough for the artefact on each channel type. After visual inspection you can close the plot and click “OK” from the dialog to write the projected data.

11. Open Calculate EOG projections dialog. Click "Plot events" to see how well blinks are identified from the automatically selected EOG channel and after that press Compute.

12. Do the same as in (10) but with "Apply EOG projections" dialog. Now data will have the projection vectors embedded and the data will be transformed accordingly when further analysis is done.

Create epoch collection for ERF inspection

13. Now look at the top of window. There are tabs called "Preprocessing", "Epoching", "Averaging", "Spectral analysis" and "Source analysis". We will now proceed to epoching so select the "Epoching" tab.

14. Select "Create new collection.." to generate a new set of epochs.

15. Let's first create a collection for left-ear auditory responses. For this, the trigger that was used to label left auditory sounds during the experiment was 1. So set the collection name to be something like "LeftAuditory”, and set Event ID to be 1. You could also change epoch length and rejection criteria here. Click "Add to list". Then press "Create epochs."

16. Let's create a collection for right-ear auditory responses. Name should be "RightAuditory" and event ID should be set to 2. By clicking the collection you can see some information about it on the box below and on the box on the right.

Visualize Evokeds

17. We can now proceed to "Averaging" tab. Our purpose is to create a plot where evoked potentials of left auditory and right auditory responses are plotted on top of each other to see if they differ.

18. Select both RightAuditory and LeftAuditory entries from the collections box and click "Create evoked dataset".

19. Entry called "RightAuditory-LeftAuditory_evoked.fif" should appear on the right. Select it and click "Visualize selected dataset" below. Topographically arranged plot of evoked responses appears.

20. You can try clicking some of the sensors to take a closer look.

21. Close all the opened plots. If you wanted to do statistical analysis outside of Meggie, you can use "Save evoked data". "Show evoked stats" dialog can calculate and save some of the usually interesting statistics for different sensor groups and types.

Plot time-frequency representations (TFRs)

22. Let's select the "Spectral analysis" tab now. We will take a quick look at the TFR's of one of the conditions. Note, that you can also compute power spectra on this tab.

23. Select LeftAuditory and click TFR topology. There are quite many settings that can be tweaked, but we'll just change the minimum and maximum frequencies for now. You can take a closer look on some sensors by clicking them.

Do source-level analysis

Coming up sometime!