Our group seeks to elucidate how large populations of neurons process information, from encoding sensory stimuli to guiding motor actions. Most neurophysiological studies to date involve studying one neuron at a time. Although one neuron can be informative about the sensory stimulus or motor action, it often doesn't tell the full story. While this provides the motivation for looking across a neural population, the heterogeneity of the activity of different neurons can be baffling (e.g., see Churchland and Shenoy, 2007 and Machens et al., 2010).
We have two major aims:
(1) To develop and apply novel signal processing and machine learning
algorithms to explain the high-dimensional structure and timecourse of
neural population activity.
(2) To apply this knowledge to the design of next-generation
biomedical devices that interface with large populations of neurons.
The work is at the intersection of signal processing / machine learning, biomedical engineering, and basic neuroscience.