Berlin 2015 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
BP: Fachverband Biologische Physik
BP 2: Neurophysics I
BP 2.12: Talk
Monday, March 16, 2015, 12:30–12:45, H 1058
The effects of positive interspike interval correlations on neuronal information filtering properties. — •Sven Blankenburg1,2 and Benjamin Lindner1,2 — 1Department of Physics Humboldt University Berlin, Germany — 2Bernstein Center for Computational Neuroscience Berlin, Germany
Neurons encode time-dependent stimuli in sequences of action potentials (spike trains). The neuron can encode either preferentially slow components of the stimulus (low-pass filter), high frequency components (high-pass filter), or a specific frequency band (band-pass filter) depending on the characteristics of the stimulus as well as the intrinsic neuronal dynamics. The standard integrate-and-fire model acts as a low-pass filter of information, irrespective of the firing regime (regular, Poisson-like, or bursting) of the neuron. However, experiments reveal band-pass filtering in some neurons, requiring theoretical explanation. In this study, intrinsic stochastic adaptation is identified as a possible mechanism. For our analytically tractable model system we show that positive ISI correlations, such as those caused by stochastic adaptation can result in a pronounced band-pass filtering of information for time-dependent stimuli.