SSVEP and alpha generation

dandeliondandelion Canada
edited May 2021 in General Discussion

Hello all!
I have purchased the All-in-One Biosensing R&D Bundle and I have mixed feelings now. On the one hand, I absolutely love the ease and the flexibility of the system, especially the ability to try different electrodes. On the other hand, I fail to replicate the results I used to have on a more expensive system. I wanted to test SNR and get some initial impression of the system so I decided to run a couple of simple EEG paradigms. I started with SSVEP. I published a couple of papers using this methodology and I have a number of stimulations, which used to produce good results. They were Matlab based programs. In addition, I created a python based stimulation and recorded it as a movie and gif files. All of them have sine modulation and some of them feature a flashing square, whereas others, images of faces and animals. Then I used all there sets of electrodes (wet, headband and headset) and I could not see any SSVEP response, although blinks in time domain and 60 Hz in frequency domain were quite prominent. I tried to run it on my household members and did not have any success either. Initially I used earlobe references then I tried to plug Cz as SRB and nothing happened. My electrode placements correspond roughly to O1, O2, PO7. PO8 locations. Then I tried to enhance alpha by closing my eyes. I could not do even that. At this point I am slightly confused and I would like to understand if do anything wrong. I followed the tutorials in setting the system, but I could not find any good tutorials for analysis in python. I saw just the one with EEGLAB. I also read a couple of threads around SSVEP here with @wjcroft. But I still have a couple of questions:
1. I wonder if anyone has encountered an MNE tutorial for the analysis of bdf files.
2. Is there any step-by-step description of the suggested EEG analysis for SSVEP?
3. Is there a requirement to do any type of scaling due to the full band raw data type?
4. Assuming that I do something wrong and it's not a problem of the system lacking sensitivity, what would be the step to identify where I go wrong?
5. Can it be due to the failure of quality issue of some of the components and if yes, how should I test it?

Any other suggestions are welcome.
Thanks!

Comments

  • wjcroftwjcroft Mount Shasta, CA
    edited May 2021

    Dandelion, hi.

    If you are generating the SSVEP stimulation, with flashing monitor screen animations -- this is highly dependent on a number of OS and monitor related technical issues. You might get some insights from the tips documented by the MindAffect project. They do not use SSVEP, but instead cVEP (code based VEP). But their tips are still relevant.

    https://mindaffect-bci.readthedocs.io/en/latest/installation.html?highlight=vsync#framerate-check [one example]

    https://mindaffect-bci.readthedocs.io/en/latest/stimulus_time-lock_test_tutorial.html [another]

    I published a couple of papers using this methodology and I have a number of stimulations, which used to produce good results.

    Are you using the EXACT same hardware, OS, monitor as the previous stimulation? Changing any of the OS, monitor, stim program, etc, can produce poor timing.

    Alpha generation varies between individuals. So if you are not seeing an increase in the 10 Hz FFT widget, OpenBCI_GUI, with yourself -- then try some friends as well. If you are not getting the 10 Hz eyes closed increase, with multiple individuals, then I would suspect something is wrong with your setup.

    William

Sign In or Register to comment.