[Eeglablist] CTF sensor coordinates and dipfit approach

[Baris Demiral]

Hi guys,

I have 275 channel CTF data. I was able to import it into EEGLAB,
selected out the MEG sensors on the scalp only only, and applied PCA,
reduced the data to 64 components, then found 64 ICs.

Well, I have some questions to those who have experience in CTF data
analysis for the next steps:
My goal is to find the dipoles from those ICs with dipfit. I can see
the gradient/sensor coordinates when I look at the 3D locations of the
sensors; +x is the nose. So far so good.

Questions:

1) The sensor coordinates are already corrected/oriented to the
fiducials recorded at the beginning of the run, is this correct?

2) The 2D scalp views of the ICs correctly represent the head
orientation, right? Nose is pointing to the +X direction -> up? Do I
have to rotate it?

3) Is there a document / web source comparing the way the EEG versus
MEG eye and ekg artifacts look alike/different on the scalp
topography? Is there a toolbox using ICA to reject the MEG artifacts
automatically (I think ADJUST and others only correct  EEG artifacts,
not MEG artifacts, is this right?).

4) When I followed the dipfit stages defined in the web tutorial
(coarse fit etc.) it asks for subject mri co-registered with the CTF
file, this has to be in .mri format. Actually I have AFNI formatted
anatomical file where I marked the fiducials. Can I convert it back to
.mri data format, or is there a way that I can use this AFNI file
instead in DIPFIT? Any suggestions?


Thanks,
Baris