search

grig95 / ZuCo2.0-processing

1 stars 0 forks source link

Valid EEG channel mismatch #2

Open grig95 opened 2 months ago

grig95 commented 2 months ago

Upon cross-referencing the channel map in the manual of the EEG system used for data acquisition (GSN-HydroCel-128 v1.0, found here: https://www.documents.philips.com/assets/20180705/6f388e7ade4d41e38ad5a91401755b6f.pdf) and the map of used channels provided in the paper, a total of 24 missing channels were found, instead of the 23 stated in the paper. See eeg_plotter.MISSING_CHANNELS for the full list. Both channel maps can be found in the channel_maps folder.

A side effect of this is that currently the last channel in the dataset is never mapped to any channel of the original 128, thus never influencing the plots (or anything based on them) in any way.

Moreover, it has been assumed, though it was never clearly stated, that channels 0-104 in the dataset correspond the the used channels in ascending order.

grig95 commented 2 months ago

Note: once this is fixed remove the if statement in eeg_plotter.channel_fill()

grig95 commented 2 months ago

This may not be a problem since it seems that eeg channel 104 is mostly or entirely 0.