Time-Frequency Coding via Wavelets Analysis for Bursting Spikes of Neuron Firing

Bursting (spikes) shows very complex dynamics of firing activities of the neurons. The aim of this study is to investigate the coding details in time-frequency domain for bursting spikes using wavelet analysis, which will give a combinative neural code from both time and frequency domains. Data in this study is two types of the spikes: bursting and spiking simulatied from Chay model via 4-order R-K algorithm under different tasking conditions. The interspike intervals (ISI) are obtained through the spike series used for the neural coding. Mexican Hat (´mexh´) wavelet is selected as mother wavelet used in this study. Neural coding of bursting under a 5-scale decomposition is performed and compared to that from temporal coding. The results of this study have demonstrated that time-frequency coding might be a more useful approach, and provide more details to neural coding for the bursting spikes than rate coding and temporal coding.