Welcome

DIGIBRAIN is a DIGITEO project coordinated by J.-D. Muller (CEA-LIST) and M. Sebag (LRI). DIGIBRAIN will investigate EEG-based interfaces for neuroscience studies and Human Computer Interaction.

Research Activities and Application Fields

Electrical brain activity can be (partially) recorded using electroencephalography (EEG). Nowadays, EEG measurement systems are mainly used in the context of epilepsy and Brain-Computer Interfaces (or more generally EEG-based interaction). DIGIBRAIN will investigate those two application fields.

Nonlinear spatial filtering in Brain-Computer Interfaces

Spatial filtering is a necessary tool to recover task-specific brain activities from the general measurements recorded by EEG from the scalp. DIGIBRAIN will investigate the use of nonlinear spatial filtering methods to improve tasks discrimination.

Unsupervised clustering in Epilepsy

Nonlinear learning methods are intended to be used for epileptic seizure prediction.

Active learning for high-level cognitive functions regulation

Brain-computer interfaces and more generally brain-machine interaction (such as neurofeedback, computer-assisted brain regulation) usually involves two learning processes: the computer tries to learn how the subject is performing some specific tasks (this part has been well studied in the BCI literature), and the subject learn how to interact with the computer using feedback in a trial and error process (this side is still obscure). DIGIBRAIN will investigate these complementary sides using an active learning framework.

Partners

CEA (Commissariat à l'Energie Atomique)

INRIA (Institut National de Recherche en Informatique et en Automatique)

LRI (Laboratoire de Recherche en Informatique)

Neurospin

External Collaborations

Epilepsy EEG Database

Database recorded at the Epilepsy Center of the University Hospital of Freiburg, Germany.