Is this the right way to preprocess eeg-data on matlab and eeglab?

Question

KaPa 2017-3-1

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/327480-is-this-the-right-way-to-preprocess-eeg-data-on-matlab-and-eeglab

Hi everybody, I actually want to analyze my recorded eeg data (recorded in hdf5 file). First I imported the hdf5 file to Matlab and preprocessed this data (detrending, filtering). After that I opened this data on eeglab to allocate the Channel locations (11 Channels) of the electrodes and run the ICA to reject artifacts. Some information of the recording and data: The experiment startet on two participants simultaneously with a 2 min resting state, followed by a roleplay between the participants and ended again with a 2 min resting state. The data contains 1 epoch (0-721.207 sec) - its about 12 minutes and so far non epoch. The sampling rate was 256 Hz. First af all I want to analyze the resting state data. So the next step was to "import event info" by a .txt file where I wrote down the type (resting state 1 and 2), the latency and duration of this both resting data. It does work, because eeglab shows me now 2 events. Now my idea was to "select data using events" - there I chose my resting state 1 and set the "time litmits" from 0 to 115 (because of approx. first 2 minutes resting state). It leads to a new dataset containing 1 event with 1 epoch from 0-115 sec. Is this correct so far? Now I wanted to segment this data into epochs of 2 sec each, make the FFT for average power spectra and calculate first the absolute power for each electrode on each frequency (alpha, beta, gamma) and second the relative power with asymmetry index. Can I do the following steps on eeglab oder should I convert the .set data to Matlab again? If yes, how can I do this?

Thank you in advance for your help Katharina