search

MAMEM / eeg-processing-toolbox

Matlab code for proccesing EEG signals.
http://www.mamem.eu/
Apache License 2.0
74 stars 36 forks source link

How do you extract the Stimulus Frequncies from the data sets #54

Open Gargablegar opened 8 years ago

Gargablegar commented 8 years ago

I have looked at the datasets, and I want to use this data as extra data in experiments I am running which use a novel model based classification approach to SSVEPs in BCIs.

To make use of this data I need to have a value at which the stimulus is presented. In your data gathering method you indicate that the stimulus was randomised. I cant seem to see anywhere where the actual Frequency of the stimulus is record.

Any help with this would be great :)

Gargablegar commented 8 years ago

I see there is a Session.m file which segments the data according to experiment. A useful output would be something like an array of [Frequency presented, Start Sample, End Sample]

Could you perhaps point me to where you assign Frequency from the DIN data? For datasets I and II.

Gargablegar commented 8 years ago

I have used: sess = eegtoolkit.util.Session; sess.loadAll(1);

To extract this data, i will follow the comment i found in the code that says "%dont ask" :p

liarosge commented 8 years ago

Hi,

Yes you can use the method "split" of class "Session.m" to segment the Dataset I&II into trials of 5 sec duration which correspond to the stimulation period. Just be careful, there is a small difference between datasets I & II.

In the first dataset the stimulation frequency is calculated from the DIN_1 data (see "freqs" variable inside split).

In the second dataset there is a "labels" variable inside each .mat file which corresponds to the sequence of the stimuli. 1=12Hz 2=10Hz 3=8.57Hz 4=7.5Hz 5=6.66Hz

Gargablegar commented 8 years ago

I would just like to say I am using this toolbox and the PSDA classification portion as a benchmark comparison to the performance of my model based classification method.