Aller au contenuAller au menuAller à la recherche

Dans cette rubrique

Séminaire - Byron Yu Brain-computer interfaces for basic science

Abstract :

 Brain-computer interfaces (BCI) translate neural activity into movements of a computer cursor or robotic limb.  BCIs are known for their ability to assist paralyzed patients.  A lesser known, but increasingly important, use of BCIs is their ability to further our basic scientific understanding of brain function.  In particular, BCIs are providing insights into the neural mechanisms underlying sensorimotor control that are currently difficult to obtain using limb movements.  I will demonstrate this advantage of BCI using two studies.  The first involves identifying a network-level explanation for why learning some tasks is easier than others.  The second involves the use of internal models identified from neural population activity to explain why subjects make movement errors. These findings deepen our understanding of how neurons interact during learning, an suggest ways to accelerate learning of everyday skills.

Selected publications

“Stimulus-driven population activity patterns in macaque primary visual cortex”
by B. R. Cowley, M. A. Smith, A. Kohn, B. M. Yu.
PLoS Comput Biol, 12(12): e1005185, 2016.

“Scaling properties of dimensionality reduction for neural populations and network models”
by R. C. Williamson, B. R. Cowley, A. Litwin-Kumar, B. Doiron, A. Kohn, M. A. Smith*, B. M. Yu*.
PLoS Comput Biol, 12(12): e1005141, 2016.

“Fault tolerance in the brain”
by B. M. Yu.
Nature, vol. 532, 2016, pp. 449-450.

Scientific focus :

Byron Yu received the B.S. degree in Electrical Engineering and Computer Sciences from the University of California, Berkeley in 2001. He received the M.S. and Ph.D. degrees in Electrical Engineering in 2003 and 2007, respectively, from Stanford University. From 2007 to 2009, he was a postdoctoral fellow jointly in Electrical Engineering and Neuroscience at Stanford University and at the Gatsby Computational Neuroscience Unit, University College London. He then joined the faculty of Carnegie Mellon University in 2010, where he is an Associate Professor in Electrical & Computer Engineering and Biomedical Engineering and the Gerard G. Elia Career Development Professor.  He is broadly interested in how large populations of neurons process information, from encoding sensory stimuli to driving motor actions.  His group develops statistical algorithms for studying large-scale neural recordings (basic science) and for brain-computer interfaces (biomedical engineering).