EEG analysis question

[Psychology 76]

Hi all,

So for part of my honors thesis I took a baseline EEG reading of my subjects. I'm looking at asymmetry between frontal areas and I think I figured out how to run the proper analysis in biopac.

The problem is that this is going to take a long long time to do (60-70 hours on top of classes, grad school stuff, GRE retake ect.) Is there anyway to run a macro or some sort of program that does the analysis for me? I imagine there has to be someway in our technological age. The analyses I'm doing are pretty strait forward (log, FFT, spectral analysis) but I'm doing it for each 4 second chuck and averaging them which is why it takes so long.

Is a way to write a program or use something like MATLAB to do this for me?
Anyone have any experience with this?

Thank you!!

---

Members don't see this ad.

 

[Ollie123]

Trying to learn more about spectral analysis myself - we primarily do ERP so I am just starting to read up on this and get familiar with the EEG side of things. There is absolutely a way to make macros like this in matlab, though if you haven't done much matlab programming before you could probably spend 60-70 hours just learning Matlab well enough to write something like that (possibly longer). Potentially less if you have a solid programming background, or if you have a script you can modify rather than starting from scratch. That said, its probably a much better investment of those 60-70 hours than doing rote, repetitive work manually. If there are other researchers at your school who do EEG work, or collaborators of your investigator, you might be able to get your hands on someone else's script.

Wish I had something I could send you, but I have 2-3 other papers I need to get analyzed and out before I can get to the EEG one! We don't seem to have a strong technical psychophys/stats presence on the board when I have asked questions related to psychophys in the past so I'm not sure what kind of response you will get, but hopefully someone has done something like this.

PS - Also just wanted to say props for taking this on as an undergrad thesis. Far more complex and technically challenging than some dissertations I've seen...stuff like this tends to be "scary" for a lot of people in the field.

 

[Psychology 76]

Trying to learn more about spectral analysis myself - we primarily do ERP so I am just starting to read up on this and get familiar with the EEG side of things. There is absolutely a way to make macros like this in matlab, though if you haven't done much matlab programming before you could probably spend 60-70 hours just learning Matlab well enough to write something like that (possibly longer). Potentially less if you have a solid programming background, or if you have a script you can modify rather than starting from scratch. That said, its probably a much better investment of those 60-70 hours than doing rote, repetitive work manually. If there are other researchers at your school who do EEG work, or collaborators of your investigator, you might be able to get your hands on someone else's script.

Wish I had something I could send you, but I have 2-3 other papers I need to get analyzed and out before I can get to the EEG one! We don't seem to have a strong technical psychophys/stats presence on the board when I have asked questions related to psychophys in the past so I'm not sure what kind of response you will get, but hopefully someone has done something like this.

PS - Also just wanted to say props for taking this on as an undergrad thesis. Far more complex and technically challenging than some dissertations I've seen...stuff like this tends to be "scary" for a lot of people in the field.

Thank you for your reply! I don't have much programming experience (though I'm really trying to learn some) but another person at my school is doing a similar analysis for her thesis and she does have some programming experience. Maybe together we can come up with something but I'm glad you said there is a way to do it because we were wondering if it was even possible. I go to a very small school so there really isn't anyone who could help us with this particular analysis but we are asking around in different departments.

Good luck with your papers and thank you for the kind words about my thesis. I really enjoy the quantitative and technical parts of psychology/neuroscience I only wish I would have gotten into it sooner so I don't have to keep teaching myself basic stuff!

 

[Ollie123]

Best of luck. By the way, what you refer to as "basic stuff" I'd wager 95% of faculty lack even a basic understanding of. I'm continually amazed that even leaders in the field frequently have absolutely no idea how any of this works. Its definitely got a steeper learning curve and as you are discovering, publication can be a lot slower because data processing alone can take as much or more time than other studies from start to finish. That said, I think psychologists reluctance to learn things that are more complicated and technical (whether out of fear, or incompetence) is a significant barrier to the scientific progression of the field.

Though I will say I think its far worse when people "try" to do that research without actually learning how, and completely botch it up, often in ways that you won't even realize if you don't know to look. I always wonder how much published research found results due to incorrect scoring templates, improper data reduction procedures, etc....

Feel free to PM me if you have further questions or want to discuss. This is the kind of stuff that I LOVE...clinical work? Pshaw....study design, methodology, and analytic techniques are where its at. Throw in some theory, and I'd be perfectly happy never seeing clients again (though I don't like the idea of completing losing touch and ascending into the ivory tower never to return) if I could do this stuff all day long.

 

[cheeze]

Some EEG programs let u write scripts I believe

 

[Ollie123]

Haven't used net station, but I believe it is exclusive to EGI systems...not sure how well it works with data from other systems (of course, if you are using EGI, than great!).

Though as an aside....booo windowed amplitude
I kid (somewhat), we are just big proponents of the PCA approach here.

As an aside - one option might be to download EEGLab if you haven't tried it yet. I have it on my computer but haven't sat down to learn it yet. It runs off Matlab and may help streamline things without requiring quite as much in the way of scripting.

 

[Psychology 76]

Haven't used net station, but I believe it is exclusive to EGI systems...not sure how well it works with data from other systems (of course, if you are using EGI, than great!).

Though as an aside....booo windowed amplitude
I kid (somewhat), we are just big proponents of the PCA approach here.

As an aside - one option might be to download EEGLab if you haven't tried it yet. I have it on my computer but haven't sat down to learn it yet. It runs off Matlab and may help streamline things without requiring quite as much in the way of scripting.

So far no dice with this... Got EEGLab to run but for some reason it can't read my matlab files.

Does anyone know where I can find more information about this? Literally, no one at my school can help me with is kind a downside at being at such a small school. Is there any kind of forum for EEG analysis? Mathworks seems a bit too general..

 

[Ollie123]

EEGLab has a great website with wiki. They also have a listserv that seems pretty responsive (too responsive - I actually resigned from it because there was too much activity flooding my inbox). Actually, if you get any book recommendations for the intro, please either post or PM me because I would be curious too.

If you are into spur-of-the-moment and have money, they also have a workshop that starts in 2 days in La Jolla