Predicting neural activity spike by spike: The power (and limits) of simple neuron models

Wulfram Gerstner

Ecole Polytechnique Fédérale de Lausanne, School of Computer and Communication Sciences,
Computational Neuroscience Laboratory, Lausanne, Switzerland

We used simple neuron models of different types to predict neuronal activity during intracellular current injection. Having previously used standard integrate-and-fire neurons with adaptation, and spike response models with sliding threshold, we now show directly from experiments that a simple neuron model should feature an exponential spike initation mechanism combined with refractory effects. The quality of the spike-by-spike prediction is scaled by the intrinsic noisiness of the neuron.

[1] R. Jolivet, A. Rauch, H.-R. Lüscher and W. Gerstner (2006) Predicting spike timing of neocortical pyramidal neurons by simple threshold models Journal of Computational Neuroscience 21:35-49

[2] C. Clopath, R. Jolivet, A. Rauch, H.-R. Lüscher and W. Gerstner (2007) Predicting neuronal activity with simple models of the threshold type: Adaptive Exponential Integrate-and-Fire model with two compartments Neurocomputing 70:1668-1673,

[3] L. Badel et al.,(2007), submitted.

[4] R. Jolivet et al., (2007) submitted

Back