Study of Bursty Neuronal Models

Biswas, Rohan (2016) Study of Bursty Neuronal Models. Masters thesis, Indian Institute of Science Education and Research Kolkata.

PDF (MS dissertation of Rohan Biswas (11MS022)) Rohan-Biswas_11MS022.pdf - Submitted Version Restricted to Repository staff only Download (3MB)

Abstract

Understanding the functioning of brain still remains a very intriguing area of research. One of the ways to understand it is through study of mathematical neuron models. From the data generated through experiments, various models have been proposed which on simulation, gives spiking activity very similar to that shown by real neurons. Burst firing is a very interesting type of neuron spiking which is believed to contain more information than normal single spike trains. We have also seen neurons to adapt to a given stimulus. We wanted to observe that a network containing both bursty and adaptive neuron can show signs of adaptation. We start by studying and simulating the famous Hodgkin Huxley model which is the most biologically plausible model. But its computational complexity compels us to study a simpler model by Izhikevich. This model enables us to study networks containing excitatory and inhibitory neurons for both regular spiking and bursting neurons. Another important aspect of spike trains is adaptation which is studied through a similar model called AdExIF model. This enables us to observe existence of adaptation in networks containing both excitatory and adaptive neuron.

Actions (login required)

View Item