I am doing some analysis about the resting state EEG.My data is about 3min for each subject, and I want to extract the epoch about 2s and calculate the power for each subject.
Thanks
Hi
Your question is too general. you should study EEGLAB TUTORIAL.
Good luck
https://sccn.ucsd.edu/wiki/EEGLAB_Wiki
Hi,
Mike Cohen's lectures have also helped me understand the general picture, although the eeglab tutorial is really well done.
http://www.mikexcohen.com/lectures.html
Good luck
hi Wenwen Chang
you might need to read a bit about eeglab tutorial, normally and in general the pre-processing of data include :
1- referencing your data to a channel or averaged channels
2- filtering: as you need to filter unwanted frequencies
3- reject some artifacts in your data such as eye blinks, you can do that by visually checking your data or use some plugin such as MARA
4- epoch your data into segments
5- you can now filter you data to specific frequency bands of interest. then you can check the signal power at these bands.
i hope this help.
regard
Atef
Search "Makoto's processing pipeline", that's pretty much best practice.
Thanks for all your kind answers.
Actually, I have read the tutorial of the EEGLAB, and in the tutorial, it is mainly about ERP which have event or trigger in the data, but for the EEG data for patient, it do not have the event or triggers, so I am not sure how to do the epoch about the EEG data.
And for the ICA, it is better for continuous data or for the epochs?
Generally you want to do ICA on epoched data, so that you're not decomposing as much non-task related activity.
Hi Wenwen,
normally it is not 100% mandatory to have the information about the times of the events to carry out preprocessing in EEGLAB. If your data are 3 minutes length, you can epoch every 2 seconds without problem. Of course, you need to remove abnormal epochs in order to do a good ICA decomposition for artifacts rejection..
@Alexander Cerquera,@all
Dear all,
Thanks for your kindly suggestions. I almost understand how to do it. For my data, my analysis include:
1.Input data, visual inspection, do the band filter (1Hz to 50 Hz) and reject the abnormal data and the bad channels.
2. channel location
3. averaged reference
4.for EEG data, I need to insert event/trigger to do the epoch, using MATLAB script,
5. do the epoch and basline correlation
6.visual inspection and reject abnormal epoch
7.Do Ica and SARICA
8. save the data, and if necessary, do the average about the epochs & filter
So this is what I did for my data, but I am not sure is there any problems.
I tend to run ICA, then reject epochs in which there is highly correlated activity between ICs (i.e. poor decomposition), and then run ICA again. You could justify your way though.
I'm totaly new to eeg analysis and was wondering what features are normally extracted when comparing between long term conditions (i.e sitting 10 minutes with eys closed, 10 minutes with eys open, and 10 minutes listening to specific music types.
Also- what is the ICA used for?
Hello, I am newbie in ERP signals. I am confused regarding the steps needed to do to extract ERP components. My experiment is on visual stimulus and collecting ERP data. Once i obtain my data what are the steps i need to do ? Which is the best tool to do so ? Any suggestions would help.
I want to know about the flow before I jump into the study. Thank you
Matthew Moore Wenwen Chang
you must follow Ali Motie Nasrabadi ,he is one of the greatest master to signal processing all around the world.
Hello everybody!
What if we have a 0.30 seconds length of edf file, do we need to extract events and epochs?
- Badges
- Science topic