Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 56234 (05)
- University: Sharif University of Technology
- Department: Electrical Engineering
- Advisor(s): Vosoughi Vahdat, Bijan
- Abstract:
- The retina is a light-sensitive layer of tissue at the back of the eyeball wall and is the initial stage of the brain's visual processing stages. This tissue contains more than 60 types of neurons placed in different neuronal layers. All the visual information received from the surrounding environment passes the retina and is encoded into neural signals. Therefore, studying the retina's functionality and bio-inspired modeling of its neurons and their connections are essential in recognizing visual dysfunctions caused by problems in the retina. Also, this modeling can help researchers in proposing solutions for building retinal prosthetics and, in general, performing simulations in the studies on this tissue. In this study, bio-inspired models of the primary neurons composing the retina, based on Hodgkin-Huxley equations, and the pathways built with them are proposed. Object-oriented implementation of these models makes them scalable and allows the user to develop larger models with various connections between cells. Moreover, employing simple methods like forward-Euler and fourth-order Runge-Kutta to solve the differential equations makes the model optimized in simulation time and computational cost. The proposed models have been tested with different inputs, and connections between the cells and the similarity between the simulation and real-cell results have been investigated. In addition, the equations have been modified to reconstruct the physiological findings better. Finally, a model of Horizontal cells is proposed and analyzed, and a solution for making a proper connection between these cells and the rod photoreceptors to form center-surround receptive fields is proposed.The models were validated based on two criteria: first, the ability of the models in encoding different input stimuli, and second, the similarity between models' functionality and physiological findings of related studies. Therefore, it is first investigated if the model can transfer the variabilities in the input to the output. Second, experiments similar to the ones in the physiological studies were done with the model, and the results were compared
- Keywords:
- Retina ; Neural Modeling ; Hodgkin-Hoxley Equations ; Differential Equation ; Bioinspired Modeling ; Ionic Current Modeling
Digital Object List
-
محتواي کتاب
- view
Bookmark