Talk by Simon Vogt
Institut für Signalverarbeitung, Universität zu Lübeck
- 26 Nov 2012 15:00
- 26 Nov 2012 16:30
- INM-6, Bldg. 15.22, Seminar Room 3009, 1. OG
A unifying perspective on opposing neuromodulatry effects during synaptic transmission in D1-dominant MSN neurons
In order to give a better explanation for neurological diseases of our brain's dopamine (DA) processing areas such as the basal ganglia and the many side effects that today's clinical treatments cause, we should aim to understand these networks' activity dynamics, learning paradigms, and single-spike timing features from an information processing perspective. By increasing our knowledge of the neural spike code to a depth where we can understand and truly decode multisite electrophysiological recordings of spiking networks, we will be able to devise more informed treatments of many neurological disorders in the future. For example, the exact effect that dopamine has on synaptic transmission in a spiking network is still largely unknown, and many biological studies have arrived at inconclusive or opposing conclusions.
While the application of (dopaminergic) reinforcement into a plastic neural network by three-factor learning rules can help to explain the influences in training outcome that varying activity of dopaminergic cells in the SNc/VTA seem to have, this broad multiplication is biologically questionable and does not account for instant dopamine-dependent changes of network activity during glutamatergic inputs, as the fast timescale of activity changes can not be the result of slow plasticity processes (e.g. R-STDP).
In my talk, I will present a unifying view on how opposing observations of neuromodulation may affect a spiking neural network, and suggest possible directions for further experimental curiosity. Instead of modelling changes of network activity as the result of slow DA-dependent weight changes, I show how DA-modulated synaptic efficacy may lead to both instant changes in postsynaptic activity and influence overall learning outcome.