AUTHOR=Lankarany Milad , Heiss Jaime E. , Lampl Ilan , Toyoizumi Taro 

TITLE=Simultaneous Bayesian Estimation of Excitatory and Inhibitory Synaptic Conductances by Exploiting Multiple Recorded Trials

JOURNAL=Frontiers in Computational Neuroscience

VOLUME=Volume 10 - 2016

YEAR=2016

URL=https://www.frontiersin.org/journals/computational-neuroscience/articles/10.3389/fncom.2016.00110

DOI=10.3389/fncom.2016.00110

ISSN=1662-5188

ABSTRACT=Advanced statistical methods have enabled trial-by-trial inference of the underlying excitatory and inhibitory synaptic conductances of the membrane potential recordings. Simultaneous inference of both excitatory and inhibitory conductances sheds light into the neural circuits underlying the neural activity and advances our understanding on neural information processing. Conventional Bayesian methods can infer excitatory and inhibitory synaptic conductances based on a single trial of observed membrane potential. However, if multiple recorded trials are available, this typically leads to suboptimal estimation because they neglect common statistics (of synaptic inputs) across trials. Here, we establish a new expectation maximization (EM) algorithm that improves such single-trial Bayesian methods by exploiting multiple recorded trials to extract common synaptic input statistics across the trials. In this paper, the proposed EM algorithm is embedded in parallel Kalman filters (KFs) and Particle filters (PFs) for multiple recorded trials to integrate their outputs to iteratively update the common synaptic input statistics. These statistics are then used to infer the excitatory and inhibitory synaptic conductances of individual trials. We demonstrate the superior performance of these multiple-trial Kalman Filter (MtKF) and Particle Filter (MtPF) methods relative to the corresponding single-trial methods. While relative estimation error of excitatory and inhibitory conductances is known to depend on the level of current injection into a cell, our numerical simulations using MtKF show that both excitatory and inhibitory condutances are reliably inferred using an optimal level of current injection. Finally, we validate the robustness and applicability of our technique through simulation studies and apply the MtKF algorithm to in-vivo data recorded from rat barrel cortex.