Hi,
I´m trying the raw file review on .edf files exported from brainAnalyser, but I get quite different signal morphology on certain electrodes. I try to view the file with eeglab and get quite the same as brainAnalyzer view. Please find attached picture about the three views (same time): brain analyzer, brainstorm and eeglab. I used a short eeg for the example available here :
http://dl.free.fr/qiQfNLV1A.
For example, the beginning of MAS1 channel is clearly different in brainstorm than in the other views. Even if the global amplitude scale is not exactly the same, there is differences in term of amplitude proportions : see for example FP1 amplitude compare to other channels.
Do you think it´s only a view matter or it could affect following treatment ? For example blink detection and ssp removal ?
Thanks,
Alexandra
Hi Alexandra,
On top of the Brainstorm window, in the toolbar, there is an “AVG REF” button. When selected, as it is in your case, it shows values that are re-referenced to average reference.
At each time point, the mean of all the electrodes is subtracted from each channel. This is why when is a strong contamination on one channel you see that it projects on all the other channels.
Uncheck this button and you would get displays that are much closer.
Another parameter that might change the display from an application to the other: the online frequency filters that are applied.
To avoid having slow drifts in the traces (that we are usually not interested in), you can apply a high-pass filter at a very low frequency (for instance 0.1Hz)
By default, Brainstorm doesn’t use any filter, but for instance Cartool uses by default a high-pass filter. This is something that you want to check on both sides.
To apply an online frequency filter to review recordings: use the Filter tab in the main Brainstorm window.
Note that you should exclude the EOG and all the other channels that are not really brain EEG from the group of EEG sensors.
To do so: right-click on the channel file > Edit channel file, then set the types of the channels to something else than EEG (you can type anything you want).
Then close the window and answer yes when prompted for saving.
Cheers,
Francois
I forgot to add that the channel location are not yet recognize, only the channel name. could come from this.
Moreover, when oppening with eeglab, I get a warning :
´Automated Overflowdetection not supported for EDF and BDF data, because Physical Max/Min values of EDF/BDF data are not necessarily defining the dynamic range.´
The position of the electrodes are not saved in the EDF file format.
You need to import them from another file.
The warning in EEGLAB is related with a check that we do not perform in Brainstorm.