Thesis defense Jörg Horschig (Donders Series 185)

8 June 2015

Promotor: Prof.dr. O. Jensen, copromotor: dr M. Bonnefond

Flexible control and training of posterior alpha-band oscillations

There is increasing evidence that ongoing neuronal oscillations can be directly related to cognitive performance in humans. For example, the performance in visual attention tasks is strongly related to the distribution of posterior alpha-band oscillations, which allow inhibiting task irrelevant information. In daily life attentional demands have to be flexibly adjusted to the environment. Improving the ability to flexiblyadjust attention to environmental demands is therefore crucial. Here, I first demonstrate that the statistical properties of the environment have consequences for how individuals adjust their posterior alpha activity; supposedly this reflects the environment-dependent allocation of attention. Further I show that some participants were better at adjusting to the environment than others and that this trait was reflected by the ability to modulate the alpha activity. Next I ask if the ability to modulate the posterior alpha activity can be trained and to what extend this training has consequences for behavior. To this end I have developed an MEG based brain-computer interface (BCI) controlled by posterior alpha activity. Participants could gain control of this BCI by allocating their spatial attention within a single session, and became better in following sessions. In a subsequent experiment, this setup is used to train the posterior alpha activity. As a consequence of the training, behavioral performance in a visuospatial attention task was modulated. Finally, I investigate the role of the frontostriatal system in the control of oscillatory brain activity, and show how alpha oscillations are involved frontostriatal communication. In sum I show in this thesis that posterior alpha oscillations can flexibly be adjusted to current attentional demands. Further this ability can be improved by training posterior alpha oscillations using brain-computer interfacing. This suggests that posterior alpha oscillations play a causal role for the allocation of spatial attention and that BCI training of neuronal oscillations can be used to augment human cognition.