[Eeglablist] (no subject)

[Tarik S Bel-Bahar]

Hi Sam, notes below, best wishes.

*****************************************************
One can do these steps and more in eeglab.
There are extensive learning resources (video, ppt, tutorial, all googlable),
and you can do most steps via gui.

For newcomers to eeglab, makoto's extensive recommendations are useful, see:
https://sccn.ucsd.edu/wiki/Makoto%27s_preprocessing_pipeline

Consider reviewing the methods from a few published high quality
articles for deciding about epochs/averaging of spectra, easy to find
via google scholar. For time-frequency for beginners check out Mike X
Cohen's excellent book on TimeFrequency Analyses
including his online powerpoints and videos.

If you are interested in QEEG, your best bet is to go with published
methods from most reputable journals as you can find.
Power ratios are used in various research areas with EEG, google
scholar is great for searching for good examples of current methods.
Of course, general caveat about possible quality/validity issues with
a sparse consumer EEG system.

Regarding your data length, it should be closer to 2-3 minutes
(somewhat of a minimum length for resting EEG conditions)
Epochs are usually broken in to ~2-3 second segments (depends on
lowest frequency you want to estimate).
If you need to work with only 30 sec of data, try first to find
similar published methods that you can emulate.

If you google eeglablist+ a topic of your choice you can see many
similar questions and extensive answers.
For example, googling "eeglab compute power spectrum" gave this as one
of the following links:
Chapter 03: Plotting Channel Spectra and Maps - SCCN
https://sccn.ucsd.edu/wiki/Chapter_03:_Plotting_Channel_Spectra_and_Maps

You may also be interested in exploring the wide array of EEG analysis
softwares summarized here, dated but useful:
https://www.hindawi.com/journals/cin/si/138743/

Of possible interest:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3391960/


On Fri, Mar 1, 2019 at 1:26 PM Sam van Bohemen <samvanbohemen at gmail.com> wrote:
>
> I have been using an EMOTIV Insight to record my EEG and I have been using MATLAB to analyse the data. I was wondering if you had any advice in regards to selecting epochs for analysis. I am able to identify some artefacts (eye blinks, movmement) and remove these sections of the data. For example if I have a recording that is 30 seconds long (3840 samples, sample rate of 128Hz) is there a standard method as to how many epochs I should break the data into? I have been attempting to compute some qEEG measures such a the delta/alpha ratio, relative alpha power and Qslowing. should I be calculating these measures for each epoch and then averaging them for all the epochs in the the recording? I have been using a highpass filter at 0.5Hz and a lowpass filter at 35Hz. I have also been applying a Hann window before performing the FFT and before calculating any qEEG measures.
>
> I have not been using EEGLAB, instead I have been writing my own scripts to analyse the data.
>
> Do you recommend using EEGLAB instead?
>
> Thanks
> Sam
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to eeglablist-request at sccn.ucsd.edu