Neuronal response latency estimation in presence of a background signal
Sprache des Vortragstitels:
12 International Neural Coding Workshop
Sprache des Tagungstitel:
Neuronal response latency is usually vaguely defined as the delay between the stimulus onset and the beginning of the response. It contains important information for the understanding of the temporal code. For this reason, the detection of the response latency has been extensively studied in the last twenty years, yielding different estimation methods. If the response can only be observed on top of an indistinguishable background signal (in the form of ongoing spontaneous firing), the estimation of the time delay can be highly unreliable, unless the background signal is accounted for in the analysis. Here we propose different parametric and non-parametric methods to investigate neuronal response latency based on detection of spikes evoked by the stimulation using interspike intervals and spike times. In particular, investigation from the first-spike latency in presence of excitatory inputs and/or inhibitory inputs is presented. Poisson process, integrate-and-fire model (Wiener process) or Leaky integrate-and-fire model (Ornstein-Uhlenbeck) are considered for modeling the single neuron firing mechanisms, and the proposed methods are illustrated on both simulated and real data.