SSVEP on openBCI

SeventeenChenSeventeenChen Shenzhen
edited May 2019 in Software
Hello, we tried to build a BCI with openBCI to observe SSVEP, the details are follwing:
  1.  We set six electrodes on the sclap, Oz, P3, P4, Pz, FCz(GND), AFz(Ref).
  2.  The stimuli we used are psychtool in matlab, which is a blinking square with diffenent frequencies on LCD screen, we have tried 7.5Hz, 8.75Hz,10Hz,12Hz and 15Hz, but we can't observe any  spectral response with peaks at the individual frequencies. 
  3.  During experiments, we tried to record 2s to 4s data, but we can't observe any spectral response with peaks, either. 
  4.  We have tried to import raw data into matlab, then filter the data,butter lowpass filter and notch filter(50Hz), fft, but we can't observe anything useful in the graph.    
If there are someone have tried this experiments, welcome to discuss with us about more details. 

Comments

  • wjcroftwjcroft Mount Shasta, CA
    Chen, hi.

    The search button in lower right column has many more threads on ssvep.


    In particular this recent thread, mentions the possibility that generating your blinking square from Matlab, might introduce substantial jitter into the resulting EEG response. This is because Matlab and laptop operating systems in general, are not designed for precisely timed realtime performance. Particularly with a heavyweight (interpreted?) language like Matlab. 

    Another factor to consider is that square wave stimulus in SSVEP, might result in additional frequency components besides the fundamental. Some sound-light machines use sine wave amplitude modulation of the LED lights, to only generate the desired fundamental frequency.


    The most success with SSVEP has been achieved with bright LED(s), driven by a dedicated processor with crystal controlled timing. The above thread also mentions Raspberry Pi, which has a DMA output facility.

    Regards,

    William

  • BillhBillh
    I worry your P3 and P4 electrodes are too far forward to pick up good visual cortex signals, though Oz should be okay.. Have you tried electrodes at O1 and O2 instead of the P ones yet? 

    I assume you are looking at fft bins that correspond to the frequency of the visual stimulus. Is the screen close enough to the subject to fill their center of vision?

  • SeventeenChenSeventeenChen Shenzhen
    edited May 2019
    William，hi，
    I'm pleasant to receive your detailed and abundant answer. The most important is the stimulus in SSVEP are better to use LEDS with crystal. 
    The other question is the gel electrodes. Are they the golde cup electrodes with Ten20 Conductive paste(sorry,we only have them).
    We have been using spike units in ULTRACORTEX MARK IV to record EEG.   
      
    Regards,
    Seventeen Chen
  • wjcroftwjcroft Mount Shasta, CA
    Definitely recommend cups with paste, your EEG signal quality will increase. Passive dry electrodes are not as accurate as active dry electrodes. But active dry systems cost in excess of $10K.
  • SeventeenChenSeventeenChen Shenzhen
    Billh, hi,
    Thank you very much for your comment. I will try the O1, O2 at once. The suggesstion is very great! If you have more details about the experiment, welcome to comment and contact with us.
    Sincerely,
    Seventeen Chen
  • wjcroftwjcroft Mount Shasta, CA
    Be sure you read Chip Audette's SSVEP post on his EEGHacker blog, link on other thread I referenced.
  • SeventeenChenSeventeenChen Shenzhen
    William, hi,
    Thanks for your reply. I will try cups with paste. If you have more details and suggesstions, welcome to comment and contact with us.

    Regards,
    Seventeen Chen


  • SeventeenChenSeventeenChen Shenzhen
    William, hi, 
    I have read the blog on the EEGHacker blog about use SSVEP to control the Hex Bug and Shark attack. I use the same sets in his Hex Bug experiment, but we can't view  spectral response with peaks, of course the SNR is not enough. So we are confused.
    Another question: active dry electrods ?   

    Regards,

    Chen 
  • wjcroftwjcroft Mount Shasta, CA
    https://www.google.com/search?q=active+dry+eeg+electrode

    Some manufacturers are listed there. Active electrodes have a tiny preamplifier physically BEHIND each sensor on the scalp. This is costly, such headsets or caps are many thousands of dollars.

    The paper listed at the top might also be relevant, though not covering the passive dry case,


  • SeventeenChenSeventeenChen Shenzhen
    William,hi,
    Thank you very much！It‘ll be helpful to our experiment.

    Regards,

    Chen
  • wjcroftwjcroft Mount Shasta, CA
    Interesting 2011 paper contrasting the different forms (amplitude modulation) of the light stimulation used in SSVEP: sine, square, triangle waves.

  • retiututretiutut Louisiana, USA
    SSVEP Widget is in currently in development for the OpenBCI_GUI!!!
Sign In or Register to comment.

© Vanilla Keystone Theme 2026