I am analyzing EEG data from a study concerned with fatigue in pretest-posttest arrangement, expecting higher fatigue in the posttest. I used FFT, smoothed the spectrogram and looked at the alpha band and some relevant ratios, like '(theta + alpha)/beta'. What I often find is overall increased power in the whole spectrogram in the posttest (so alpha power is indeed higher, but delta, theta, beta and gamma are as well). Can this be interpreted somehow or is this just an artifact, possibly caused by long duration of the experiment (which lasts more than one hour)?
Tomáš - I think the first relevant question would be whether this could be an artifact due to your processing of the EEG data, e.g., you might have thrown out more/less components in the ICA (if you have done one) in the post-test. That's quite easy to track by comparing the number of components retained or looking at the coefficient of variation of your spectral power over time (which is an estimate of signal-to-noise ratio). Once you have excluded this type of analysis artifact, I think increased alpha/theta power etc. has been linked to attention, fatigue etc. before (a very quick literature search found me this, for example:
Cajochen, C., Brunner, D. P., Krauchi, K., Graw, P., & Wirz-Justice, A. (1995). Power density in theta/alpha frequencies of the waking EEG progressively increases during sustained wakefulness. Sleep, 18(10), 890-894. ). So my (also quick-and-dirty) intuition is that your findings probably are plausible and interpretable.
However, I wonder why you would consider it an artifact if your power changes would be caused by time-on-task - isn't that what fatigue is all about, growing exhausted because of time-on-task?
Cheers, Marius
Thank you very much, Marius Keute, I'll answer your points one by one.
I have not conducted ICA, I have so far looked at individual electrodes (I use BIOPAC MP150 which only allows me to record limited number of electrodes at once). The signal being processed in FFT is therefore pre-processed recording from one electrode only.
You are correct that alpha/theta power is indeed linked to fatigue. What I view as suspicious is that there is increase in power also in beta band, which should be lower under fatigue. I enclose the spectrogram in question to this reply. So my question is: what might cause the overall higher power in all frequency bands?
And to your question why I would consider it an artifact - I am in fact just a novice EEG analyst, but I was thinking that maybe there are changes in the impedance of the electrodes over time, causing the recorded power to increase.
To reply to Moses P Moorthy
- thank you, but I think this was not the answer to my question.- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping