Gatsby Computational Neuroscience Unit


Ueli Rutishauser


Wednesday 19th July 2017


Time: 4pm


Ground Floor Seminar Room

25 Howland Street, London, W1T 4JG


Probing the mechanisms of memory at the single-neuron level in humans


Much of what we know about how neuronal circuits enable behaviors such as learning and decision making stems from studying the response of individual neurons in animal models. While powerful, this leaves us unable to approach many important questions for lack of similar data in humans. We take advantage of neurosurgical procedures to record at single-cell resolution in behaving humans. I will review results on the relationship between neuronal activity, plasticity, and memory derived from single-neuron recordings in the human hippocampus, amygdala, and substantia nigra. I will describe a putative circuit composed of two functional cell types, visually-and memory selective neurons, whose interaction is mediated by theta oscillations. Visually-selective neurons are tuned to high-level concepts, are sensitive to attention, and their activity forms attractors through persistent activity over several seconds while stimuli are held in working memory. Memory-selective neurons, on the other hand, signal whether a stimulus is novel or familiar, a property that changes after a single learning trial. Together, these results begin to provide a circuit-level understanding of human memory.