Noise enhancement of pattern discrimination in temporal coding

Juliana M. de Assis, Francisco M. Assis

DOI: 10.14209/sbrt.2024.1571032315

Evento: XLII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2024)

Keywords: Temporal coding Entropy Interspike Interval Hodgkin-Huxley

Abstract

Temporal coding is one mechanism of neuronal information processing. In this paper, we investigated whether there is a significant contribution of extrinsic noise in temporal coding, quantitatively. In order to do so, we simulated different patterns through white noise current, and we used the celebrated Hodgkin-Huxley model to obtain interspike intervals. Then, we fitted gamma distribution to the interspike interval histograms of each pattern and we estimated parameters of shape and rate in each case. These estimated parameters enabled the calculation of information measures, such as differential entropy. These information measures provided us evidence of the improved extrinsic noise contribution to the discrimination of patterns in neuronal temporal coding, when compared to the contribution of mean value of synaptic current.

Download