A nonlinear identification method for modelling the three-dimensional structure of velocity storage in the vestibulo-ocular reflex (VOR).

Item

Title
A nonlinear identification method for modelling the three-dimensional structure of velocity storage in the vestibulo-ocular reflex (VOR).
Identifier
AAI9108180
identifier
9108180
Creator
Sturm, Deborah Dorit.
Contributor
Adviser: Theodore Raphan
Date
1990
Language
English
Publisher
City University of New York.
Subject
Computer Science | Biology, Neuroscience | Engineering, Biomedical
Abstract
Velocity storage is present in vertical and roll components of compensatory eye movements if the head is oriented appropriately with regard to gravity (Raphan & Cohen, 1983; Matsuo & Cohen, 1984; Raphan & Cohen, 1987). To explain this, it has been postulated that gravity orients the eigenvectors of the three dimensional representation of velocity storage towards the space vertical (Raphan & Cohen, 1987; Sturm & Raphan, 1988). The thesis devised a computational procedure to identify the parameters of the system matrix from OKAN data. This procedure should help test this hypothesis in behavioral paradigms.;Velocity storage which is expressed by OKAN (Raphan et al, 1979), has been represented as a dynamical system, x{dollar}\sp\prime{dollar} = Hx, where x is a three dimensional vector representing the state of the system and H is the matrix containing the parameters that govern its dynamical behavior. Because of the assumption of linearity, the dynamical system was solved to give the pitch, roll, and yaw eye velocities in terms of the eigenvalues and eigenvectors of the system matrix. These parameters appear nonlinearly in the solution as a function of time. A nonlinear least squares fitting procedure (Marquardt, 1963) was adapted which compares data from cross-coupling experiments to the model outputs in order to identify the eigenvalues and eigenvectors of the system matrix. The eigenvalues and eigenvectors were determined using OKAN data from a monkey for a few tilt angles. When the parameters were set appropriately in the model, it closely predicted the cross-coupling effects.;Thus, a computational procedure has been developed which uses nonlinear identification techniques to identify parameters of a model of the three dimensional representation of velocity storage. This should be a useful tool in further theoretical and experimental studies on the vestibular ocular reflex.
Type
dissertation
Source
PQT Legacy CUNY.xlsx
degree
Ph.D.
Item sets
CUNY Legacy ETDs