Saving Data

Last updated: 18-June-2024

We really will only need a couple pieces of raw data saved. These are:

1. Training data (sometimes called Calibration data) where we have a ground “Truth” of what the person was attempting. This will require (1) EEG data, (2) python “markers” from FlickTok when events are started/stopped published to LSL,
2. Activity data. This will again need (1) EEG data from the headset, (2) Markers from Python when a video changes between “available” and “not-available” states (i.e. can be interacted with), (3) Outputs from the BCI prediction, timestamped, (4) An indication from the participant that they are beginning to try to attempt movement

We could save other information like video length, but I don’t think it is necessary right now.