Predictive Coding and Spike Timing Dependent Plasticity in Primary Visual Cortex

Abstract

The aim of the present thesis is to elucidate whether the brain uses the statistical regularities of its environment to facilitate faster information processing and to investigate the mechanisms it uses to adapt to changes in these regularities. To this end, psychophysical studies in human and non-human primate are combined with electrophysiological single cell recordings from primary visual cortex (V1) of the macaque monkey. In the first part, we show that the flash-lag effect—a visual illusion in which the position of a predictably moving object is misjudged relative to that of a briefly flashed object—can also be observed in monkeys. By studying the responses of single neurons to moving and flashed bars, a potential mechanism put forward to explain the flash-lag effect, the differential latencies hypothesis, is rejected. In the second part, a potential mechanism underlying the dynamic reorganization of neural circuits in response to changing statistical regularities of their environment is examined. This mechanism, spike timing dependent plasticity, has been shown to induce “predictive shifts” of neural population responses after repeated presentation of the same stimulus sequences. We demonstrate that motion conditioning can induce such predictive shifts in the visual system of awake monkeys. In contrast to previous findings, however, they are specific to the exact nature of the stimulus (moving object) rather than being manifested in general receptive field shifts. This could be a sign of a principled mechanism by which circuits in early visual areas can undergo changes as a result of learning without at the same time getting disturbed under conditions …