Rate coding versus temporal order coding: a theoretical approach
Rate coding versus temporal order coding: a theoretical approach
复制标题
DOI：
10.1016/s0303-2647(98)00050-1
复制
发表时间：
1998-09-01
期刊：
影响因子：
1.6
通讯作者：
Thorpe, S
中科院分区：
文献类型：
--
作者：
Gautrais, J;Thorpe, S
神经元以放电频率编码的形式传递信息的观点几乎是普遍的。然而，我们认为，至少在某些情况下，基于速率编码的编码策略的效率是令人惊讶的差。一个简单的数学分析表明，由于尖峰信号产生的随机性，即使是可靠地传输最简单的信号也需要：（1）与感觉处理速度不相容的过长的观察时间;或者（2）与感觉通路所施加的解剖约束不相容的过多的冗余神经元。我们认为，这样的问题可能会避免使用替代时间代码，依赖于整个人口的传入神经元的放电。(C)1998爱思唯尔科学爱尔兰有限公司保留所有权利。
The belief that neurones transmit information in the form of a firing rate code is almost universal. However, we argue that at least in some situations, the efficiency of a coding strategy based on rate coding is surprisingly poor. A simple mathematical analysis reveals that, due to the stochastic nature of spike generation, even transmitting the simplest signals reliably would require either: (1) excessively long observation periods incompatible with the speed of sensory processing or (2) excessively large numbers of redundant neurones, incompatible with the anatomical constraints imposed by sensory pathways. We argue that such problems may be avoided by using alternative temporal codes which rely on the asynchrony of firing across a population of afferent neurones. (C) 1998 Elsevier Science Ireland Ltd. All rights reserved.