A Simulation Study on Neuronal Ensemble Entropy Coding under Stimulus Condition

The aim of this study is to investigate a new method of coding neural ensemble activity via information theory and show the validity of this method in coding different stimuli (a pulse current and Gaussian white noise). Experimental data used in this paper is a simulation spatio-temporal spike of 120 neurons in hippocampus CA3 computed from small-world neural network model. We quantify neural ensemble temporal patterns via methods of informatics: estimate entropy values of neural population on temporal resolution (¿=1 ms) and word length (L=8) in a selected window (100ms) with a moving step (size of 1/4 window), and calculate the amount of mutual information between pre and post different stimuli. Then "ensemble threshold" is applied to obtain neuronal ensemble. The results indicate that difference of spike trains between pre and post stimulus is statistically significant (p<0.01). Besides, neuronal ensembles of neural firing to the pulse current and Gaussian white noise are also different. Our results demonstrate that external stimuli are encoded via neural ensemble activity and the validity of the ensemble entropy coding method in coding different stimuli has been proved.