Filtering EEG data - How to remove noise

BeatriceBBeatriceB Germany
in Cyton
Dear All,

Currently I am using Cyton board with gold cup electrodes in my EEG Neurofeedback study. I have only 8 channels (TP8, C4, C3, P4, PZ, P3, TP7 and in the ear EEG electrode) and 2 references (mastoids: TP9 and TP10). Since I am not running any ICA (because of few channel locations), I am using four filters: notch (50 Hz), low pass (60 Hz), high pass (1 Hz), and band pass for alpha (8-12 Hz). Do you think this might work in terms of artifact attenuations? I apply these filters using the EEGLAB filter function "pop_eegfiltnew".

Another thing: right now, I am not cleaning the signal from possible eye blink artifacts (I don't have frontal channel locations and I instruct participants when to blink and when not). Do you think this can be ok?

Let me know as soon as you can, please.

Thank you for your time, help and patience :)

Best,
Beatrice

Comments

  • BillhBillh
    edited March 2019
    Rule of thumb is apply filters and look at the resulting plots. When they show clean beta and when present (eyes shut, back of head leads) clean alpha you are done. If the record as plotted goes flat, change the scaling til you see the beta again--if you cannot do so you are filtering too much. 

    You should be able to see eye blinks in leads located on the front of the head as waves that are slower than alpha, but where you say put the leads -- TP8 and P4 -- you might not see the eye blinks due to the leads' distance from the eyes.

  • wjcroftwjcroft Mount Shasta, CA
    Beatrice and Bill, thanks.

    Here is a previous thread mentioning the SARA algorithm for automated removal of artifacts, including eye blink,


    Appendix 1 has the detailed description,


    Regards,

    William

  • BeatriceBBeatriceB Germany
    Hi Bill and William,

    Thank you so much for your useful suggestions! I will try to have a look to this S.A.R.A algorithm and I hope to have good results.

    And yes Bill, when I filter my data (offline analysis), I scroll the data and in the plot I can see clear alpha waves... The only thing now, was to find a way to clean better my data from possible artifacts... 

    Thank you again for your time, help and patience. I really appreciate it! :)

    Best,
    Beatrice 
  • BeatriceBBeatriceB Germany
    Hi Jackmartin2,

    Thank you for your help, I will take into consideration your suggestions.

    Best,
    Beatrice 
Sign In or Register to comment.

© Vanilla Keystone Theme 2026