UNIVERSITY OF HERTFORDSHIRE COMPUTER SCIENCE RESEARCH COLLOQUIUM presents "Inhibitory plasticity and optimal encoding at the cerebellum input layer" Dr. Sergio Solinas (Institute of Neuroinformatics, UZH / ETH Zurich) 21 April 2020 13:00 - 14:00 Over Zoom Abstract: The cerebellum's input layer, the granular layer (GL), receives massive excitatory projections from the somatosensory system, the proprioceptive system, the brain cortex, and the basal ganglia, along with minor contributions from the other sensory systems. The GL dynamics have been extensively studied, starting from Albus, Marr, Ito, and Eccles' seminal work. Nonetheless, the GL adaptive mechanisms affecting its dynamical encoding to accommodate changes yielded by growth, injuries, or ageing remain unclear. Starting from electrophysiological experiments, we investigated the role of a heterosynaptic plasticity rule of the inhibitory synapses in the evolution of the GL response to stimuli by building a network model accounting for this new plasticity mechanism. The heterosynaptic synaptic plasticity adapts the efficacy of inhibitory synapses to keep the excitatory neurons, projecting to the next layer of the cerebellum, responsive to incoming stimuli and preventing their firing activity from saturation. We envisage that our modelling work will reduce the neuronal numbers of the GL, thus facilitating the GL network transfer to a low-power, mixed analogue-digital chip able to emulate the neuronal activity in real-time.