Dia: Dimecres, 20 de desembre de 2017
Lloc: Aula T2 (2n pis), Facultat de Matemàtiques i Informàtica, UB.
A càrrec de: Toni Guillamon, Universitat Politécnica de Catalunya
Títol: Mathematical strategies to estimate synaptic conductances
Resum: We will present different approaches to estimate synaptic conductances and discern between excitatory and inhibitory inputs using only the information of the time course of the membrane potential of a neuron. In neuroscience, this information can provide insights on the excitation/inhibition balance in parts of the brain. From a mathematical point of view, the purpose of this problem is to quantify the input that a multidimensional externally forced system is receiving from the only information of a single variable. It results, then, in an inverse problem and, as other estimation problems, it brings up interesting challenges for many branches of mathematics: statistics, stochastic processes, dynamical systems, Bayesian inference, optimization,... In this talk, we will focus on aspects closer to dynamical systems, but we will also try to give an overview of other methods. In particular, we will discuss on the influence of active ionic channels, and the necessity of both inferring conductances from single trials and providing model-free methods. We will show our proposals to overcome nonlinear effects in the subthreshold regime, as well as a proof of concept for the spiking regime. The contents of this lecture are based on different collaborations with Rune W. Berg, Pau Closas, Susanne Ditlevsen, Rafel Prohens, Antonio E. Teruel and Cati Vich.
Last updated: Fri Dec 15 11:04:12 2017