EEG source localisation questions

[Ajay Halai]

Dear FT users,

I am a 1st year PhD student and new to FIELDTRIP, and any help will be very
welcome. I have managed to run simple ERP analyses on a language experiment.
I am now hoping to localise the peaks (P100 and N400) but am coming across
some problems.

I will try my best to explain briefly my current method of analyses and
would greatly appreciate any comments or suggests, as I fear I am making a
mistake somewhere.

The problem I encounter is that the values returned for the avg.pow are the
same for each condition. Naturally, I made sure that I specified different
conditions at the ft_sourceanalysis (cfg, data) stage, but I did specify
different conditions. I also checked the avg.noise, and was surprised to see
these values were the same as the avg.pow, which leads me to think I have
done something wrong or am failing to interpret this. Furthermore, I tried
to then use loreta, and selected the timelock data, with the same cfg but
get an error (Undefined function or method 'loreta' for input arguments of
type 'struct') and wonder if I need to specify other cfg.

I used the 'lcmv' method. projectnoise 'yes', rawtrial 'no', lambda = 0, and
using the vol, grid and elec files from the created leadfield.

Any suggestions on what I may be doing wrong would be greatly appreciated.
Additionally, I have provided some more details about the stages before
this, which may point to the problem.

Best wishes 
Ajay Halai

analysis details:

I have 7 conditions in all, but take 2 conditions for 1 subject as an
example. I have two types of speech, which show significant differences at
P100.  I used a bandpass filter (1-40hz), padding (0.1), baseline correction
(-0.1 0), detrend and average EEG reference at the preprocessing stage. I
removed artifacts by first removing components using the 'fastica' method,
and then a visual rejection. Following this, I specified the time of
interest limit to the P100, and used the timelockanalysis. I have not used
keeptrial or keepindividual cfg, although I don't think this is the cause of
the problem.

I do not have individual MRIs, therefore I have used a template from the
MNI. I have applied a segmentation (using volsegment). I specified the elec
positions (64 sensors), and prepared leadfields using reducerank 2,
threshold 0.1, smooth 5, units 'mm'

___________________________________________________________________
Neuroscience and Aphasia Research Unit (NARU)
School of Psychological Sciences (Zochonis Building)
University of Manchester 
Brunswick Street Manchester
M13 9PL
UK

----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/neuroimaging/fieldtrip.