I am a student of MSc. Bionics, doing my research in analysis of EEG signals. I am new to this area of research. i would be really grateful to your valuable comments and advice.
The EEG signals i have are that of children performing CPT task, to analyse their attention.
Hi Alwin,
It depends on what you are looking for as your feature of interest.
In genera,l if you are looking for frequency domain features which are lower than 62.5 Hz ( the Nyquist frequency with 125 Hy sampling rate) it should be alright. However, the safest approach is to perform downsampling after filtering and preprocessing.
Hope it helps.
Hi Sam,
Thank you for the reply.
I am looking for activities in 4 bands alpha, beta, theta and delta. And my frequency domain of interest is upto 45Hz. I did the downsampling after filtering the data. So according to your view, my data should be okay for further analysis. Thanks a lot for the help.
Hi Allwin,
The value you mentioned should be ok. However, my advice is to downsample the data first to around 128 Hz up to 250 Hz (if possible), then apply high-pass filter at 1 Hz. After that, you can continue preprocessing your data considering, e.g., removing bad channels, re-referencing the data to average, etc.
HTH.
Samer
Hi Samer,
Thank you for your reply. My raw EEG data is 500Hz. I filtered the data first and then downsampled it to 125Hz. I was just wondering if the downsampled data would give any wrong results during the further preprocessing steps. So your reply confirms that my data is good for the preprocessing. Thank you for the help once again!
Hi Nikola,
Sorry for my late reply. But why would you assume the data would be full of artefacts?
- Badges
- Science topic
- Similar topics
- Statistics
- Survey