自发和诱发神经放电序列的内在时序特性

运用Fano因子分析法，考察豚鼠听神经单纤维的自发放电序列、小鼠海马CAl区神经元的自发放电序列以及蟾蜍缝匠肌肌梭传入神经的诱发放电序列的时序特性，结果显示自发和诱发放电时间序列均存在Fano因子随计算窗口时间的增大而持续增长的特点，而原始数据的随机重排替代数据则没有这一特性，说明这些神经放电时间序列与一般的随机点过程不同，存在长时程相关性，在时序上具有某种结构特征。进一步的研究表明，这一时序结构特征可以通过将随机产生的一维正态分布序列数据，与神经放电时间序列数据进行跟随排序后而体现，提示这一特征与放电间隔的分布特点无关。

INHERENT FEATURES OF THE TIME SERIES IN SPONTANEOUS AND EVOKED NEURAL SPIKE TRAINS

Fano factor analysis was used to characterize time series from spontaneous spike trains of individual auditory nerve fibers in guinea pigs, spontaneous spike trains of hippocampal neurons in mice, and evoked spike trains of individual afferent nerves of muscle spindles from the sartorius muscle in toads. We found that Fano factors increased persistently with increasing window sizes in all three types of neural spike trains, but not in the shuffled surrogate data. Our results suggested that these spike time series had some temporal structure, which was different from general stochastic point processes. Further study demonstrated that this temporal structure can also be observed in sorted stochastic Gaussian-distributed surrogate data with spike sequences.