Our inferences about brain mechanisms underlying perception rely on whether it is possible for the brain to 'reconstruct' a stimulus from the information contained in the spike trains from many neurons. How the brain actually accomplishes this reconstruction remains largely unknown. Oscillatory and synchronized activities in the brain of mammals have been correlated with distinct behavioural states or the execution of complex cognitive tasks and are proposed to participate in the 'binding' of individual features into more complex percepts. But if synchronization is indeed relevant, what senses it? In insects, oscillatory synchronized activity in the early olfactory system seems to be necessary for fine odour discrimination and enables the encoding of information about a stimulus in spike times relative to the oscillatory 'clock. Here we study the decoding of these coherent oscillatory signals. We identify a population of neurons downstream from the odour-activated, synchronized neuronal assemblies. These downstream neurons show odour responses whose specificity is degraded when their inputs are desynchronized. This degradation of selectivity consists of the appearance of responses to new odours and a loss of discrimination of spike trains evoked by different odours. Such loss of information is never observed in the upstream neurons whose activity is desynchronized. These results indicate that information encoded in time across ensembles of neurons converges onto single neurons downstream in the pathway.

The efficiency of neuronal encoding in sensory and motor systems has been proposed as a first principle governing response properties within the central nervous system. We present a continuation of a theoretical study presented by Zhang and Sejnowski, where the influence of neuronal tuning properties on encoding accuracy is analyzed using information theory. When a finite stimulus space is considered, we show that the encoding accuracy improves with narrow tuning for one- and two-dimensional stimuli. For three dimensions and higher, there is an optimal tuning width.

No abstract

No abstract