[Eeglablist] <no subject>

[James Desjardins]

Hi Guilia,

While trying to compare the results of multiple EEG systems I wrote a  
plugin to interpolate scalp channels to the coordinates found in an  
*.sfp, *.elp, etc.. coordinate file (e.g., take files recorded with a  
256 channel EGI net, a 128 channel EGI net and a custom 128 channel  
BioSemi cap and spatially resample all the files to the coordinates  
found in a standard 81 channel extended 10-20 montage coordinate  
file). This plugin is basically just a coordination of the readlocs,  
coregister and eeg_interp functions in EEGlab, but was designed to be  
easily implemented into batching scripts.

While this works well for the analysis of scalp data in study sets  
(forces every file to have an identical scalp montage) I do not know  
what the effects would be on subsequent ICA decompositions.

Let me know if this sounds like something that would be useful to you,  
if so I will add some documentation and make it available.

James Desjardins
Technician, MA Student
Department of Psychology, Behavioural Neuroscience
Cognitive and Affective Neuroscience Lab
Brock University
500 Glenridge Ave.
St. Catharines, ON, Canada
L2S 3A1
905-688-5550 x4676


Quoting Scott Makeig <smakeig at gmail.com>:

> Guilia -
>
> To combine datasets with 128 and 64 channels, construct an EEGLAB STUDY
> containing datasets with 64 channels. Does the 128-channel montage include
> the 64-channel montage as a subset? If so, the reduction from 128- to
> 64--channel datasets should be simple.
>
> If not, then you may want to spatially resample. My suggested approach for
> that would be to perform ICA (optimally, see the wiki tutorial) on each of
> the 128-channel data sets, then resample the 3-D projection of each
> independent component to the locations of the the 64 channels used in your
> other datasets. There is not an EEGLAB function to do just that, but it
> should not be difficult (for someone) to build. Then reconstitute each
> 128-channel data set by summing the back-projections of all the 128
> components (perhaps omitting selected 'noise' components) to the 64
> channels. Note that you will then no longer be able to separate all (<=) 128
> components when you decompose the new data using ICA (since the data now
> have only 64 degrees of freedom).
>
> A possibly simpler those less refinable method would be to sample the 64
> channels in the interpolated scalp maps of every raw data point - requiring
> the same scalp map resampling function.
>
> Resampling the data from 64 to 128 channels is probably less optimal (and
> less efficient). For one, it would still allow recovery of only 64
> independent components because the rank of the data matrix (>>help rank)
> would remain 64, even when 128 channels have been resampled from the scalp
> maps.
>
> Scott Makeig
>
> On Tue, Apr 5, 2011 at 8:23 AM, Giulia Righi <grighi at gmail.com> wrote:
>
>>  HI all
>>
>> I was wondering whether there is a way to combine datasets with different
>> numbers of channels (64 and 128 channel nets) to include all participants
>> within a study design to run ERSPs and ITCs?
>>
>> thank you
>>
>> giulia
>>
>>
>> _______________________________________________
>> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
>> To unsubscribe, send an empty email to
>> eeglablist-unsubscribe at sccn.ucsd.edu
>> For digest mode, send an email with the subject "set digest mime" to
>> eeglablist-request at sccn.ucsd.edu
>>
>
>
>
> --
> Scott Makeig, Research Scientist and Director, Swartz Center for
> Computational Neuroscience, Institute for Neural Computation & Adj. Prof. of
> Neurosciences, University of California San Diego, La Jolla CA 92093-0559,
> http://sccn.ucsd.edu/~scott
>