Fig. 5
From: Chisco: An EEG-based BCI dataset for decoding of imagined speech
EEG preprocessing pipeline (part (f) cites previous work35). (a) Raw data were recorded at a sampling rate of 1,000 Hz, stored in the original .edf files. (b) The raw data were then downsampled to 500 Hz. (c) The PREP algorithm was applied for bad channel detection and re-referencing. (d) Power line noise was removed, and high-pass filtering was performed. (e) The data were segmented according to the experimental paradigm. (f) Bad data spans and breaks were rejected using the Autoreject algorithm. (g) ICA algorithm was used to remove noise such as EOG and EMG.
