|search externally:||Google Scholar, Springer, CiteSeer, Microsoft Academic Search, Scirus , DBlife|
My primary research interest is developing new methods for non-invasive brain-computer interfaces (BCIs). BCIs are devices that allow subjects to communicate by intentionally modulating the electromagnetic field of the brain. While nowadays most healthy subjects are capable of operating a BCI that allows basic communication, up to date subjects in late stages of amyotrophic lateral sclerosis (ALS) have failed to communicate by means of a BCI. It is my conviction that in order to advance the current state-of-the-art in BCIs, and enable BCI-communication for subjects in late stages of ALS, we need a better understanding how different brain regions interact in order to solve specific (BCI-related) tasks.
Accordingly, my research focuses on developing and applying methods for connectivity/causal inference in neuroimaging data. Approaches I currently pursue include
- Network information theory
- Causal Bayesian networks
- Granger causality.
It is a particular concern of mine to foster exchange between neuroimaging and machine learning, e.g. by organizing the NIPS 2009 Workshop on Connectivity Inference in Neuroimaging.
I am also interested in spatial filtering for BCIs, particularly in
- Multiclass Common Spatial Patterns
- Independent Component Analysis.
An introduction to causal inference in neuroimaging
as author at BBCI Winter School on Neurotechnology, Berlin 2014,
Video Journal of Machine Learning Abstracts - Volume 4
as author at Video Journal of Machine Learning Abstracts - Volume 4,
together with: John Shawe-Taylor (editor), Alfons Juan-Císcar (editor), Samuel Kaski (editor), Davor Orlič (editor), Jan Rupnik,