To create images for statistics one would need a 2D projection of electrode locations in the 3D space. For scalp electrodes this can be done automatically but for intracranial electrodes you are likely to just get a mess. You can try the automatic route if you change the channel type of all your intracranial channels to 'EEG'. Then SPM will let you load your sfp file. But the way I did this kind of thing before is created a channel template manually by using the 'add channel' menu item and then adjusting positions by clicking on labels (so they get green) and then clicking on the desired positions. You can then save this template to use for all your recordings from that patients. Also don't forget 'Apply' and 'Save'. Also statistics on images would only make sense for grids not for depth electrodes and separate grids should be processed separately (with Bonferroni correction for multiple grids).

On 30 May 2013, at 19:33, Franz Lukas Hell wrote:

Dear Vladimir,

I want to calculate statistics (e.g. testing for the difference between two conditions). As far as I know, I need to convert my preprocessed .mat into images before doing so.

SPM uses spm_eeg_convert2scalp for that, and then I get an error stating:

Error using griddata (line 109)
X and Y must be same length as Z or the lengths
of X and Y must match the size of Z.

I just remember that calculating images from EEG data was somehow not possible if I havent specified electrode locations. So I thought, it might be as well the problem here.

But if you can tell me a way how to do statistics without electrode locations, I am fine. Any suggestions here?

Thanks, Franz

CC: [log in to unmask]
From: [log in to unmask]
Subject: Re: [SPM] importing channel locations for intractranial eeg data
Date: Thu, 30 May 2013 07:07:12 +0300
To: [log in to unmask]

Dear Franz,

Why do you need to load positions for those channels?

Vladimir

On 29 May 2013, at 19:03, Franz Lukas Hell <[log in to unmask]> wrote:

Dear All,

I am using SPM8 to analyze intracranial EEG data (some depth electrodes and an ECoG). In similar fashion to preparing an EEG dataset I want to load an iEEG locations file (I have a +.sfp document with x,y and z MNI coordinates and channel labes which I can easily transform into an other coordiante system).

Now I choose (under "Prepare) that my "Channel types" are all LFP. When I now want to go on an load a Sensors from the Sensor dropdwon menu, SPM doesnt let me. I tried a few things, but I can't figure out how to import channel locations for LFP channels.

Can anyone help me out here? Vladimir :)

Best,

Franz