Hello great coders, I am a master student in applied psychology. I am processing a piece of children's resting-state EEG data. At present, the data has been preprocessed (including rejecting bad segments, filtering, etc.) , calculate noise covariance, compute head model and compute sourses. I wonder if these processing steps is reasonable.
In addition, I am going to divide the original data into multiple 2s segments and calculate the phase-locked value. I would like to ask how to divide the resting state data into multiple segments on brianstorm?
I would appreciate your answer, thank you!
It is difficult to judge without any details.
If you followed the instructions in the introduction tutorials, you should be OK.
The main question is probably what to do with these source-level resting-state signals.
Make sure you read all these tutorials first:
- https://neuroimage.usc.edu/brainstorm/Tutorials/RestingOmega
- https://neuroimage.usc.edu/brainstorm/Tutorials/Connectivity
- https://neuroimage.usc.edu/brainstorm/Tutorials/CorticomuscularCoherence
I am going to divide the original data into multiple 2s segments and calculate the phase-locked value. I would like to ask how to divide the resting state data into multiple segments on brianstorm?
You have an option "Split in time blocks", when epoching the data:
https://neuroimage.usc.edu/brainstorm/Tutorials/Epoching
Note that if your 2s separations are random, there is no reason for which you should expect any alignment between them. Averaging, or phase-locking analyses across 2s blocks do not have any reason to produce anything meaningful.
Thank you so much for your suggestion.
My purpose is to see the functional connections under each channel (connections based on scalp electrodes, and connections based on a certain brain area such as the dlpfc seed point).
I also asked my tutor, he said that plv is the calculation of time-correlated EEG signals. So now I will consider trying other connection methods.