Hi. I have eeg data that are in microvolts from emotiv.Fs=128 aquisition time=1min. When i plot the signal with the time vector t=0:1/Fs:60 i can see it.But when i do fft and plot it with the frequency vector freq=(1:length(signal))*fs/length(signal) the plot is so small i have to zoom in to see it.How can i set the axis to microvolts to prevent this?

 채택된 답변

Star Strider
Star Strider 2016년 2월 9일

0 개 추천

You may have a significant d-c offset. Subtract the mean of the signal from the signal before you do the fft. This will have no effect on the frequency content of your signal.

댓글 수: 2

vaggelis papasot
vaggelis papasot 2016년 2월 11일
Thank you very much for your answer!I had a suspicion about the dc offset,but reading the emotiv epoc specifications pdf https://emotiv.com/product-specs/Emotiv%2520EPOC%2520Specifications%25202014.pdf it says it has 50-60hz notch filter and thought it was to cut line noise.Now plotting a channel AF3 in time domain before cutting of the dc has its first value at 500uV,then cutting dc by AF3=AF3-mean(channel1) centers the signal at 0.Also plotting the signal with dc off at frequency domain gives a clear view of the frequency plot and is more visible that the signal is filtered between 0.2-45Hz.I hope that if anyone else is having the same problem with eeg data from emotiv or emokit will benefit from this question.Thanks again for your answer!
Star Strider
Star Strider 2016년 2월 11일
As always, my pleasure!
My own experience with biomedical signal processing led me to hypothesise the d-c offset problem. Thank you for your detailed analysis.

댓글을 달려면 로그인하십시오.

카테고리

도움말 센터File Exchange에서 EEG/MEG/ECoG에 대해 자세히 알아보기

질문:

2016년 2월 9일

댓글:

2016년 2월 11일

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by