We have tried to detect blinks and artifacts in EEG signals using brainstorm which has mostly tallied with the one’s detected manually. could you please suggest how to remove these detected blinks and artifacts as it is found that hte option is disabled in brainstorm. In artifact removal it is observed that the freqquency range is from 1 to 7Hz and 40 to 240Hz are detected as artifacts. But iam not in a poisition to cut or remove the signal in the frequency range of 1 to 7Hz as my data consists of information in this particular range. The last clariffication to be made: Is it possible to convert the data back into ASCII/text once the blinks or artifacts are removed.
could you please suggest how to remove these detected blinks and artifacts as it is found that hte option is disabled in brainstorm
If the option is disabled, it is because you have already imported your data in your database. Use the menu "Review raw file" instead of "Import EEG/MEG", as instructed in the tutorials:
http://neuroimage.usc.edu/brainstorm/Tutorials/ChannelFile#Review_vs_Import
http://neuroimage.usc.edu/brainstorm/Tutorials/ArtifactsSsp
In artifact removal it is observed that the freqquency range is from 1 to 7Hz and 40 to 240Hz are detected as artifacts
These two frequency bands are the ones in which possible artifacts are detected. The possible noisier time segments are indicated by new events in the recordings:
http://neuroimage.usc.edu/brainstorm/Tutorials/BadSegments#Run_.2301
Is it possible to convert the data back into ASCII/text once the blinks or artifacts are removed.
After importing the cleaned recordings, you export them back to ASCII. Right-click on an epoch > File > Export to file.
Note that text files are never a good option for saving recordings: slower, bigger, less efficient...
Thank you sir. I have removed blinks and artifacts by using remove simultaneous option. first i tried with a minimum delay between the events of 250ms but some blinks sustained then i changed it into 8000ms. does it make any changes to data? Also iam not able to convert it in to text files. is there any reduction for data points while removing the blinks if so does it effect the signal quality?
Hello,
Thank you sir. I have removed blinks and artifacts by using remove simultaneous option. first i tried with a minimum delay between the events of 250ms but some blinks sustained then i changed it into 8000ms. does it make any changes to data?
I'm sorry, I'm not sure I understand what you tried to do here.
This process "Remove simultaneous" is simply for removing the blinks that are too close to a cardiac event, so that the two artifacts don't get mixed up when you compute the SSP decomposition. Read carefully the tutorials.
Note we do not recommend the SSP approach for cleaning EEG recordings. Maybe you should consider ICA instead (follow carefully this entire tutorial):
http://neuroimage.usc.edu/brainstorm/Tutorials/Epilepsy#Artifact_cleaning_with_ICA
Also iam not able to convert it in to text files.
If you want help here, please provide more details.
is there any reduction for data points while removing the blinks if so does it effect the signal quality?
The blinks are never "removed" from the data. The samples that you tag as "blink" are never removed.
If you mark them as bad segments, they will be ignored when your data is epoched:
http://neuroimage.usc.edu/brainstorm/Tutorials/EventMarkers#Bad_segments
http://neuroimage.usc.edu/brainstorm/Tutorials/Epoching
Cheers,
Francois
Hi,
Please refer to this topic regarding the same ìssue with “runica”: ICA issue
You can always try JADE instead.
Cheers,
Martin