Electrical and Electronics Engineering PhD Thesis Defense by Deniz Kılınç

KOÇ UNIVERSITY

GRADUATE SCHOOL OF SCIENCES & ENGINEERING

ELECTRICAL AND ELECTRONICS ENGINEERING

PHD THESIS DEFENSE BY DENİZ KILINÇ

Title: Stochastic Analysis and Simulation of Noise in Neuronal Circuits

Speaker: Deniz Kılınç

Time: May 23, 2019, 14:00

Place: SNA B153

Koç University

Rumeli Feneri Yolu

Sariyer, Istanbul

Thesis Committee Members:

Prof. Alper Demir (Advisor, Koç University)

Prof. Alper T. Erdoğan (Koç University)

Assoc. Prof. Emre Mengi (Koç University)

Asst. Prof. Funda Yıldırım (Yeditepe University)

Asst. Prof. Güneş Ünal (Boğaziçi University)

Abstract:

The brain is extremely energy efficient and remarkably robust in what it does despite the considerable variability and noise arising from the inherently stochastic mechanisms that exist in the neurons and the synapses. For example, these noise sources can significantly degrade spike timing precision, which is believed to be a key factor in sensory information encoding and computations performed by the neurons and neuronal circuits. Computational modeling is a powerful tool that can help us gain insight into neuronal mechanisms utilized by the nervous system. In addition, a deep understanding and computational design tools can help develop robust neuromorphic electronic circuits and hybrid neuroelectronic systems. In this thesis, we present a general modeling framework for biological neuronal circuits that systematically captures the nonstationary stochastic behavior of ion channels and synaptic processes. In this framework, fine-grained, discrete-state, continuous-time Markov Chain (MC) models of both ion channels and synaptic processes are treated in a unified manner. Our modeling framework features a mechanism for the automatic generation of the corresponding coarse-grained, continuous-state, continuous-time Stochastic Differential Equation (SDE) models for neuronal variability and noise. Furthermore, we repurpose non Monte Carlo noise analysis techniques, which were previously developed for analog electronic circuits, for the stochastic characterization of neuronal circuits both in time and frequency domain.  These techniques, which are orders of magnitude faster than traditional Monte Carlo type simulations, can be used to directly compute the spike timing jitter variance, power spectral densities, correlation functions, and other stochastic characterizations of neuronal circuit operation. We verify that the fast non Monte Carlo analysis methods produce results with the same accuracy as computationally expensive Monte Carlo simulations. We have implemented the proposed techniques in a prototype simulator where both biological neuronal and analog electronic circuits can be simulated together in a coupled manner. By using this simulator, we investigate several distinct neuronal circuit motifs and mechanisms such as synaptic feedback inhibition, synaptic integration, and synaptic coupling. First, we show that both the spike timing precision and the energy efficiency of a spiking neuron are improved with feedback inhibition. We unveil the underlying mechanism through which this is achieved. Second, we demonstrate that a neuron can improve on the timing precision of its synaptic inputs, coming from multiple sources, via synaptic integration: The phase of the output spikes of the integrator neuron has the same variance as that of the sample average of the phases of its inputs. Finally, we reveal that weak synaptic coupling among neurons, in a fully connected network, enables them to behave like a single neuron with a larger membrane area, resulting in an improvement in the timing precision through cooperation.