[FieldTrip] ICA and plotting

[Nathan Weisz]

hi jason,

it may be that you are rejecting too much, i.e. removing too much relevant brain activity. in general with regards to blinks and horizontal eye movements (your main sources of ocular artifacts) you should reject ~2, not (significantly) more. have you confirmed that your method is not too liberal in detecting "artefact components"?

since the time-course and topography of these artefacts are so clear, "visual inspection" may be your best friend.

good luck!
nathan


On 26.10.2011, at 16:38, Jason Chan wrote:

> Hi Everyone,
>  
> I am currently using Fieldtrip version 20111009. 
>  
> I am trying to use ICA to reject EOG artifacts.  I ran a correlation  between my EOG electrodes and MEG components to find the components that should be rejected.  This variable I called “toreject”, then used the following function:
>  
> cfg = [];
> cfg.component = toreject;
> data_reject = ft_rejectcomponent(cfg, comp);
>  
> When I try to plot my data using ft_topoplotER, the brain activity looks ‘spotty’.  However, when I add the cfg.component back into the cfg structure the activity looks correct.  I have also noticed that adding this cfg.component will change the plots after I use MEGplanar and CombinePlanar. What am I doing wrong?
>  
> Thank you in advance
> Jason
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20111026/860c9c4b/attachment-0002.html>