[FieldTrip] weird topo: due to missing sensors? and [Reference to non-existent field 'chanori']

[Stephen Whitmarsh]

Hi Vitoria,

Yes, I guess it might be the missing channels somehow.

Concerning the missing chanori; I think you throw away more than just the
data when you remove a channel, and when you restore it later on, things
like the orientation, location etc. is not in the data anymore. I would
therefor never throw away a channel, but rather fix it, to avoid these
kinds of problems. Anyway, I think you can solve it by copying the .grad
structure from an earier point (before having removed channels) into the
data after fixing.

If this doesn't solve the topo problem, have you tried dividing the
difference by the sum of the conditions or otherwise normalizing the
difference for each subject? E.g. (Condition A - B) / (Condition A + B).

Good luck!
STephen




On 19 September 2014 01:13, Vitoria Piai <v.piai.research at gmail.com> wrote:

>  Hello FT-ers,
>
> I was wondering whether anyone has an idea of what could be going on in
> this topo of the group-averaged ERF (difference between 2 conditions,
> planar gradient). At first I thought that could be muscle activity, but
> that pattern seems to hold regardless of the time window I choose (and
> besides when I plot the EMG of some facial muscles, they don't differ from
> each other...).
> What I noted now is that different sensors were excluded for different
> participants around those areas so I'm suspecting that the missing sensors
> are causing the problem (because of the way the data is interpolated for
> plotting?).
>
>
> I tried different options in ft_topoplot but all my attempts were in vain.
> Next, I tried using ft_channelrepair. I assumed that the correct order of
> processing steps would be to repair channels first, and then calculate
> planar gradients after. If I follow that procedure, I get the following
> error when running ft_megplanar (FT version 20140518):
> Reference to non-existent field 'chanori'.
>
> Error in ft_megplanar (line 242)
>   chanposnans = any(isnan(sens.chanpos(:))) || any(isnan(sens.chanori(:)));
>
> If anyone could shed some light on these issues, it would be great!
> Thanks a lot,
> Vitoria
>
> _______________________________________________
> 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/20140919/9ffeb48f/attachment-0002.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: hjahgfhc.png
Type: image/png
Size: 14934 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20140919/9ffeb48f/attachment-0002.png>