User's Guide‎ > ‎Analysis‎ > ‎

2 - Artefacts rejection & single subjects averages


When you e.g. recorded your EEG with BIOSEMI, a bdf-file (raw data) is available for each subject and session incl. also the triggers (i.e. time points of events (stimulus, response, etc.) occurrence within the EEG file)


1.       Open Cartool

2.       Tools > Averaging EEG Files

a.       Input Files: Drag & Drop all you want from Windows Explorer (only the bdf `s of ONE subject at a time)

b.       Trigger/Marker: define all you want (separating them by a space, below e.g. trigger 1 & 2 for Epoch 004)

i.      When one line is completed press “Add Edit Line to Lists”

ii.      These “Lists” can then be saved by pressing “Write Lists to File” and writing an *.csv (which can then be edited in Excel and extended like you would do e.g. in SPSS syntax scripting or for creating *.bat or *.sh files)

iii.      If lists already exist, import them by pressing “Read Lists from File”

c.       Options:

i.      Press “Triggers” when you take triggers directly from the *.bdf file

                                                ii.     Press “Markers” when you want additional triggers from an *.mrk file or a *.tva file


TIP: If you want to combine two or more triggers (e.g. in the bdf-file you have 1, 2 and 3 coded separately but now you just want them all to be named “X”) do it here! Put them all into the “Triggers/ Markers” field separated by SPACE and give the output file the name “X” then…

3.       Averaging Parameters

TIP: All parameters can be imported by dragging & dropping a .vrb-file from a previous averaging into the window!

a.       Time Interval: define pre- and post window of analysis (relative to trigger onset) in ms

b.       Set threshold in µV  (e.g. 60 or 80 µV) for automatic rejection over head and/or ocular electrodes (if measured)

c.       Choose electrodes you know to be excluded in each bdf of one subject (here e.g. B14 and C17)

d.       Filtering: set high- and low-pass filter cut-off frequencies (e.g. 0.1 HP and 40 LP and 50Hz Notch Filter)

i.      TIP: The band will automatically be adjusted as a function of the length & the sampling frequency of the epoch, just press “ok”

e.       Baseline correction: ON/OFF as you like or obliged to by your boss (eg. 50-0 means 98ms pre-stimulus baseline correction with 512Hz sampling)

f.         Don’t switch on “average reference” for single subject averages, as with average reference, bad electrodes will influence the other electrodes

g.       Output = *.eph or .ep

i.      Merging/Splitting Triggers (common average or averages subdivided by trigger label)

ii.     Data computed: Save Average and Save Standard Error (as default)

4.       Outputs files
                        a.   somehow self-explaining

TIP: Check carefully your subject’s BDF before the averaging, so that you can exclude all “bad” electrodes. This could really help to decrease the number of rejected triggers (epochs) since one bad electrode can indeed be responsible for numerous rejections, independent of eye blinks or other punctual artefacts.


Please note: The default setting is to exclude channels from visual artifact inspection but keep them in the final file.  The default is active if "write Excl. Channels" in the Options files is set. This is in general very helpful: If this box is not activated and you exclude channels, the channels will be physically removed from the averaged file ! So be careful here. If this is done by accident it may result in different numbers of electrodes for the different subjects in a group!

You can interpolate (i.e. recalculate based on surrounding channels) bad channels in a next step. If you need to reduce the total number of channels, this can be done either in the "export" function or while interpolating.