The laboratory of Neural Computation is interested in understanding how populations of neurons in the cerebral cortex encode and the exchange information. The laboratory investigates these issues by developing mathematical methods for single trial analysis of neural population responses and by developing biologically plausible neural network models of cortical function and dynamics.
Mathematical methods for the analysis of neural population responses
We develop methods, mostly but not only, based upon Shannon's Information Theory of communication for the analysis of time series of cortical recordings from multiple locations. The algorithms are designed to determine which neural oscillatory pathways or network nodes provide information relevant for perception or behaviour, what information they carry, and when and how these pathways or nodes exchange information.
Biophysically plausible models of cortical dynamics and of cortical encoding
We use models of recurrent networks of spiking neurons with the aim of understanding the dynamics of primary sensory cortices under naturalistic stimulation conditions, and to derive simple coding rules describing the transformation between the dynamics of natural stimuli and that of cortical oscillations.