These are the codes I used to preprocess the SEED VII EEG dataset (https://bcmi.sjtu.edu.cn/home/seed/seed-vii.html).
The preprocessing steps are (Step1_Preprocessing.m):
1. Bandpass filter @ 1-50 Hz (pop_eegfiltnew).
2. Rereference using common average (pop_reref).
3. Bandstop filter @ 50 Hz (CleanLine).
4. Independent Component Analysis (runica - infomax extended).
5. Dipole fitting (dipfit).
6. ICLabel.
After that, I manually reviewed all components and rejected those suspicious (Step2_ReviewICA.m). You can find the rejected components in rejected.md.
Finally I selected the data corresponding to each emotional state from the original dataset (Step3_SegmentData.m):
1. Disgust.
2. Fear.
3. Sad.
4. Neutral.
5. Happy.
6. Anger.
7. Surprise.
With the data preprocessed and segmented, I evaluated a pre-trained Neural Network to label the EEG windows with an emotion (Step4_Evaluation.m).
by: Diego Caro López, 13-Jan-2026.
Queries: dgcarolp@hotmail.com // A00833057@exatec.tec.mx