Dear colleagues
I am a postdoctoral fellow in neurophysiology with a background in psychology. I am conducting a study about the Event-related Spectral Perturbation (ERSP) of neurological patients during motor execution tasks. Everything is going well and we already have the ERSP built by MATLAB EEGLab. However, I also want to continue with an analysis of the temporal variation of the power for each frequency, that is, convert the ERSP maps into curves as in the works:
1)
Article A Comparison of Independent Event-Related Desynchronization ...
2)
Article Event-Related Desynchronization/Synchronization in Spinocere...
or even creating a 3D map:
https://www.fuw.edu.pl/~durka/papers/unification/node8.html
thank you for all your help.
best
Lucas
Hi, in my impression, you can use the newtimef() function of EEGLAB to extract the ERSP data, and the ERSP data is a 4-D matrix (with subject * nbchans * freqs * timepoints, if you cycle with the subjects). So, if you want to extract and plot the power curve, just comfim the specific frequency and channel, and average across the subjects.
Besides, if you use the STUDY module to calculate the ERSP, you can use std_erspplot() function to extract the ersp data, its also a 4-D matrix, it seems to be freqs * times * channels * subjects. And the rest of the operation is the same as above.
Dear Liu
Thank you so much for your answer.
In fact you got my point, that's exactly what I need.
However, due to my lack of knowledge of Matlab I don't understand how to use newtimef(). I tried to use some models I found but none worked. Can you help me write the script for my STUDY?
My parameters are:
sampling rate - 250hz
frequency range - 1 to 50hz
time range - -1000 to 5000ms
data file - "Test1.study"
Frequency bands - Alpha (8-12), Beta (12-30), LowBeta (12-20) and HighBeta (20-30).
I want to generate a table with all the power value for each of these bands, for each participant.
Thank you Liu
- Badges
- Science topic