Learning effects in 2D trajectory inference from low-frequency EEG signals over multiple feedback sessions

 Recent research from our group has shown that non-invasive continuous online decoding of executed movement from non-invasive low-frequency brain signals is feasible. In order to cater the setup to actual end users, we proposed a new paradigm based on attempted movement and after conducting a pilot study, we hypothesize that user control in this setup may be improved by learning over multiple sessions. Over three sessions within five days, we acquired 60-channel electroencephalographic (EEG) signals from nine able-bodied participants while having them track a moving target / trace depicted shapes on a screen. Though no global learning effect could be identified, increases in correlations between target and decoded trajectories for approximately half of the participants could be observed. Keywords Electroencephalography, trajectory decoding, learning effects.