Table of contents
4. Main points of EEG signal preprocessing and data analysis
4.1 Review of basic knowledge of EEG
4.3.3 Segmentation and baseline correction
4.3.5 Bad Guide Removal/Interpolation
4.3.6 Independent Component Analysis (ICA)
4.4 Event-related potentials (ERPs)
4.4.2 What should be reported in ERP research
4.4.3 How to present ERPs results
4. Main points of EEG signal preprocessing and data analysis
4.1 Review of basic knowledge of EEG
The research method of the action potential signal is different from the latter two.
4.2 Artifacts
This glitch-like is myoelectricity. Elimination method: average superposition.
4.3 EEG preprocessing
Basic principle: keep the original data as much as possible; relative principle, 10%.
Preprocessing process: import data->locate electrodes->remove useless electrodes->re-reference->filter->segmentation and baseline correction (save)->interpolate bad leads and remove bad segments->run ICA(save)-> Remove noise components: such as remove eye movement components (blink, eye drift) -> save data.
Online references are not very important because of heavy references.
Interpolation of bad guides: Use a certain algorithm to replace bad guides.
4.3.1 Filtering
In EEG, high-pass filtering is often used.
4.3.2 Re-referencing
Different ingredients have different effects on the reference.
The average reference should be used with caution, as it has a greater impact on the waveform.
4.3.3 Segmentation and baseline correction
After segmentation and averaging, the ERP waveform is obtained.
4.3.4 Bad segment removal
4.3.5 Bad Guide Removal/Interpolation
4.3.6 Independent Component Analysis (ICA)
ICA method for denoising.
4.4 Event-related potentials (ERPs)
4.4.1 How to get ERPs
Waves obtained from the same stimulus are superimposed on average.
4.4.2 What should be reported in ERP research
4.4.3 How to present ERPs results
4.5 Summary
The ICA method can only help us exclude typical artifacts, but cannot effectively extract EEG responses.
