If we have the Fp1 electrode (with reference to ear electrode). What kind of research can we do (with only Fp1 electrode). The frontal lobe of a brain is exposed to EOG artifacts. Where I can find information on brain waves from this area? Can we use paradigm SSVEP, P300 or ERD/ERS?
Hi Wojciech!
One electrode is not much. Did you loose all the other frontal electrodes due to EOG artifacts? Many ERP components are strongest in the frontal electrodes (like the P300), so you could do that (although more frontal electrodes would have been better). You can then argue that for P300 frontal electrodes are most important, so no other electrodes are needed.
Good luck!
Thank you Marion!
I have a very simple eeg device. It has only one electrode. It is used to a simple BCI. However, I connect this eeg to Matlab. I get raw eeg signal, which I can decompose to spectrum based on FFT. I'm looking what I can do with this.
It depends on your paradigm what you can do with it. Is it just a recording without any task or did you present the subject with some sounds/visual stimuli?
If you have no task, you can go for FFT analysis only. If you have some kind of a task, say an 'oddball'-task your best option would probably be P300. But maybe others have more ideas what to do with one electrode.
I would like to generate two state. This is not important how and which paradigmat will be used. The main issue is that I can use only Fp1 electrode and I have a raw signal. The FFT analysis can be helpful. I'm thinking about detecting relax state, but I'm not sure how do it.
Hi Wojciech! As Marion said, one electrode is not so much! But, I would like to ask you why your electrodes are placed on Fp1 and on the earlobe? Can't you move them over other scalp areas? For example if you put one of these two electrodes over the posterior regions (referred to earlobe or Cz) you could record alpha activity obtaining two states: open eyes (very low power within the alpha band), closed eyes (high power within the alpha band). Or, placing one of the electrodes over the central regions you could record an activity related to rest/movement (calculating ERD). The Fp1 electrode is always affected by eye artifacts and is not suitable for a good EEG analysis design. If you have only one channel you cannot use ICA to clean eye artifacts.
Anyway, if it is impossible to move the electrodes from Fp1-A1, then you could try to carry out a very simple "task related desynchronization" experiment, asking your subject to keep his/her gaze fixed on a point on the floor, more or less at 1-2 meters from the seat, avoiding eye movements (this is a good way to reduce eye artifacts). Then ask him/her to move the right hand for one minute and to stay relaxed for another minute, then compare the spectra of the two conditions... maybe you can find a reduction of 8-12 Hz activity during the movement.
Marco, thank your very much for your answer!
I know that one electrode, which is placed in unfortunately area Fp1, is not so much. The EEG device, which I have, is the cheapest BCI device (This is Neurosky MindWave). The electrode is placed on the stiff shoulder, and its design allows only for working in the Fp1 area. Therefore, I can't move them over other scalp areas.
In summary, the device reads the raw signal only from Fp1 area, and the reference electrode is localised on the tip of the ear.
I would like to do a kind of preliminary studies to get funds for buying a better EEG device. So far, the MindWave is connected to the Matlab environment. I used the dynamic libraries of the manufacturer. I can use the FFT to analysis the spectrum of a raw signal. I know that accuracy of this device is not high. Additionally, the FP1 area is always affected by eye artifacts. Your advice on how to minimize an importance of the eye artifacts in the signal is invaluable to me. Thank you very much.
I would like to ask you how compare in the best way spectra of two time period?
Wojciech, you can compare the two spectra in this simple way.
1. extract the power of the band of your interest, say alpha band. You will have Power_rest and Power_mov
2. normalize in this way: 100*(Power_mov - Power_rest)/Power_rest.
This is the percentage of power decrease (if negative - task related desynchronization) or increase (if positive - task related synchronization) of your "mov" condition compared to the "rest" one.
- Badges
- Science topic