[Eeglablist] about a blinking period in an EEG protocol

Ross Fulham ross.fulham at newcastle.edu.au
Sun Jun 28 17:50:53 PDT 2015

Current practice in ERP studies seems to be: (a) Provide a fixation point or other visual cues that help participants minimize eye movement (b) provide general instructions about reducing eye & body movements, but to not specifically instruct participants to inhibit eye blinks during the task, (c) apply either ICA or one of the regression based techniques (eg Semlitch) to remove eye-blink related artifact, (d) Assume residual eye movement artifact averages out during computation of averaged ERPs.

Some people include a separate recording block during which they specifically ask the participant to perform a series of (calibrated) eye blinks and vertical and horizontal eye movements. They then use this data to compute eye-blink and eye movement correction parameters. This might be of use when working with some clinical groups, e.g. Parkinson's, where eye blinks are relatively infrequent.

I have seen some researchers specifically instruct participants to blink immediately after each trial. I do not advocate this approach  as it tends to synchronize the artifact to the stimuli, and it is a secondary task that distracts the participant.

For auditory tasks, eye blinks 'tend' to be relatively random w.r.t. the task, especially for unattended tasks such as Mismatch Negativity. So blinking is less of an issue.

For visual tasks, spontaneous blinking becomes synchronized to the presentation of the stimuli. Participants suppress blinks for the 200 ms prior to an expected stimulus. Blinking recommences about 200 ms post stimulus. Consequently, participants seldom 'miss seeing' the visual stimulus because of blinking. Not sure if this behaviour has been thoroughly examined in clinical groups. Spontaneous blink rates in clinical groups is related to dopamine levels, so in schizophrenia blink rates are high, and in Parkinson's blink rates are low. Because of the blink suppression, residual blink artifact will not seriously impact early ERP components, but it 'might' have an effect on say P300 and later responses.

Not sure if blinking affects the underlying cognitive processes, but I have seen evidence that cognitive processes can alter blink rates over very short time frames.

Ross Fulham

Research Officer, Newcastle University.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20150629/aaea51ba/attachment.html>

More information about the eeglablist mailing list