I need to read instructions in the SPM Manual for
"re-referencing to average electrode" carefully since I have not
reach to 3D Source Reconstruction. But your suggestions are very
important, thank you!
I have recorded 5 subjects. After I family with the processing
steps, I will record and analysis another 5-7 subjects. We use
three kind of HydroCel Geodesic Sensor Nets size (adult small:
53.5-55cm, adult medium: 55.5-58cm, and adult large:
58.5-61.5m). They are all 256-electrode. Now, I am not sure I
have the spatial information of each and every electrode yet.
Vladimir said that SPM comes with several possible sets of
coordinates for EGI caps that are in EEGtemplates subdirectory.
So I will try to find how to identify and load it manually.
- your script for transforming NetStation EEG data to readable
SPM data
- a .pos file for example, that delivers my information of how
to build a .pos file that gives SPM coordinates about every
electrode.
Date: Thu, 26 Jan 2012 01:12:36 +0100
From:
[log in to unmask]
To:
[log in to unmask];
[log in to unmask]
Subject: Re: [SPM] EEG NatStaion (EGI) -> SPM8
Dear Deng,
again, I mail a copy to Vladimir Litvak (please read).
if you use 3D Source Reconstruction you will have to adept
your data to a brain model - either a template, which SPM
offers you, or your MRI scans of the subject.
As you go farther with 3D Source Reconstruction, SPM will ask
you for an input of coordinates of each electrode. This
information is essential!
You will have to feed SPM with the information of talairach
coordinates for each and every electrode (and the fiducials)
and if you did not measure them, I see a major problem.
You need to feed the software with information about the
location of each electrode (without this ifnromation, what
would a inverse construction with a head model do?!).
If you didn't measure the coordinates of each electrode, you
might consider using a standard set of all electrodes (i could
offer you some of my 128-electrode settings measured with our
subjects), but that would NOT be something you want to do,
because the head shapes of subjects do vary, of course.
What I want to tell you, is that if you want to use 3D Source
Reconstructuin, then you need a head model (which can be an
MRI scan or a template that SPM delivers you) and the
information of the localisation of each electrode in that
space.
If you don't have the spatial information of each and every
electrode, how could SPM build a proper model?
What I mean is: Depending on your EEG-net, electrode Nr.1
could be frontal or parietal or anywhere else. Do you know
what I mean?
Again, this E-mail was sent to Vladimir and you, and I am sure
he will respond to you (and me) if i did get anything wrong.
Tomorrow I will send to you:
- my script for transforming NetStation EEG data to readable
SPM data
- a .pos file for example, that delivers you information of
how to build a .pos file that gives SPM coordinates about
every electrode.
I wish feedback from you and Vladimir Litvak,
Greets to both of you.
Urs Bachofner
Am 26.01.2012 00:33, schrieb wangdeng:
Dear Urs,
Yes, the data is fine after I adjust the EEG lines by
using the buttons.
I would need to use Source reconstruction because I want
to compare the activities locations with fMRI data. My
study is to investigate how the brain uses information
from our senses to guide perception and action
(Multi-sensory integration). Both EEG and fMRI data will
recorded.
So I just have used Artifact removal, QRS detection, OBS,
and then save as *.raw. Now, I have converted into SPM
file format. Next, I will use SPM to analysis.
Thanks your email.
Cheers,
Deng
Date: Thu, 26 Jan 2012 00:16:25
+0100
From:
[log in to unmask]
To:
[log in to unmask]
CC:
[log in to unmask]
Subject: Re: [SPM] EEG NatStaion (EGI) -> SPM8
Hi Deng,
I respond to both you and vladimir right now.
My answers are based on my own experiences, I am sure
Vladimirs response should be weighted more.
That your data looks "strange" could simply be a visual
problem. You can adjust the EEG lines by using the
buttons above (in amplitude and in a temporal way).
The electrode locations, if you need them (if you wanna
use Source Reconstruction) can be easily read by loading
them as a ".pos" file. Which is basically a renamed .txt
file with a certain structure that SPM can read. I could
mail you one of mine so you can see how it should be
built up.
I don't know what you're planning to do with your data
and if you really need the location of all electrodes.
If you planned to use them (for example for source
reconstruction), you should have measured them with a
software and some pointing tool that gives you
coordinates. We use MrIcron in our team.
Also the fiducials, which are both anions, nasion
and(/or) inion can be measured with such a tool, if you
really need it for you analysis. These are three mark
points (= a 2D plain) that are needed so SPM can allign
the coordinates of every single electrode with the brain
structure or brain template. Again, I don't know what
you want to do with your EEG data.
For the Netstation's Waveform Tool: Everything that can
be done there can be done equally or better with SPM.
After you load your EEG file into SPM, you can filter
(highpass, lo-pass, notch (60 or 50hz, depending on
country)), remove artifacts by eye or bye SPM
automatically, etc. For these preprocessing tools I
recommend reading the SPM manual.
Now I am wondering what you want to do with your data,
because it depends on this, what preprocessing steps are
really important to you.
My experience is: Don't do too much in the NetStation
Software, just make your files work in SPM and you are
in good hands.
Best,
Urs Bachofner
Am 25.01.2012 23:35, schrieb wangdeng:
Dear Urs and Vladimir,
I really appreciate your answer!
I just used the method 1 (as Vladimir and you wrote
to me). It works! just gives warnings 'Warning:
Could not obtain electrode locations automatically.'
and 'Warning: Could not obtain fiducials
automatically. ' but the EEG data looks strange
(attached).
I will watch the video carefully.
I have tried the method 2 (write a script), but the
error message says "Out of memorry". So it would be
better if I can see your matlab script because
structure information (sampling rate, number of
channels etc) is important for analysis. So I wish
you would send me your script.
Thank you suggest me to use some steps in
NetStation's Waveform Tools. Does that means just
use 0.1-100Hz bandpass and 60Hz notch? I have used
Artifact removal, QRS detection, OBS, and then save
as *.raw.
Thank you again that you like to help me. I am alone
to process the EEG-fMRI data.
Have a good night~
Cheers,
Deng
Date: Wed, 25 Jan 2012
23:02:27 +0100
From:
[log in to unmask]
To:
[log in to unmask]
Subject: Re: [SPM] EEG NatStaion (EGI) -> SPM8
Hi Deng,
I absolutely understand your problem and had such
problems myself.
Now, to help you: I am currently not in my office
but I will be there tomorrow.
1. The first thing you can try is to use the
Convert button in SPM. As shown in the following
Video by Vladimir Litvak (he is the MASTER and
helped me a lot, too). Choosing your EEG file with
"convert", SPM can read a lot of EEG formats. It
is shown in this video of Vladimir at the time
2:34
http://www.ucl.ac.uk/stream/media/swatch?v=336ec1018a46
(the whole video is worth watching)
2. If this shouldn't work (I got to know this
option too late), maybe I can help you by showing
you how we did it:
In the SPM Manual (which you can find here:
http://www.fil.ion.ucl.ac.uk/spm/doc/manual.pdf)
you can read that the EEG file needs to have a
certain structure. Have a look at chapter 12 (page
97 and further). You will see that all the
important inputs like sampling rate, number of
channels etc need to have a distinct name.
Therefor I asked our computer guy to write me a
matlab script, that transforms all the important
information from the original EEG file to the
structure SPM will read.
I am not a programmer myself, but if you
understand a little about programming matlab
scripts, I am sure you can see the important steps
if I would send you that script. Or maybe it will
even work 1:1 without adaptions.
I have to tell you, though, that we did some
filtering steps in NetStation's Waveform Tools
before I loaded the file into SPM.
I can send you this script tomorrow, when I am in
office again, if you wish me to do so.
I would like to help you, because I also had
problems and was all alone and people helped me
until everything worked fine.
Don't worry, it CAN be done.
Greetings from Switzerland,
Urs Bachofner
Am 25.01.2012 22:33, schrieb wangdeng:
Dear Urs,
Thanks for your response!
My problem is:
I recorded EEG data using Geodesic NetStation
EGI. I will use SPM8 to process the data. But
I can't import/load EEG data because SPM needs
two kind of compatible files: *.mat and
*.data. EGI support said they don't directly
support SPM.
So could you tell me (1) what kind of file
format should be saved as, and (2) how can I
load/import EEG data using SPM?
Cheers,
Deng
Date: Wed, 25 Jan
2012 22:23:19 +0100
From:
[log in to unmask]
To:
[log in to unmask]
Subject: Re: [SPM] EEG NatStaion (EGI) ->
SPM8
Hi Deng,
my team uses Geodesic NetStation EGI and I
am processing those data in SPM. What's the
problem?
Greets,
Urs
Am 25.01.2012 22:14, schrieb wangdeng:
Dear all,
Does anyone use SPM8 to process EEG data
from NatStaion (EGI)?
My Question is how to use SPM8 to load
EEG data?
Thanks very much,
Cheers,
Deng