Chen, Xisha (2012) Pulsatile electrical stimulation of auditory nerve fibres : a modelling approach. PhD thesis, University of Warwick.

Preview

Text WRAP_THESIS_Chen_2012.pdf - Submitted Version Download (4Mb) | Preview

Request Changes to record.

Abstract

A stochastic leaky integrate-and-fire nerve model with a dynamical threshold
(LIFDT) has been derived for the neural response to sinusoidal electrical
stimulation. The LIFDT model incorporates both the refractory effects and the
accommodation effects in the interpulse interactions. In this thesis, this
phenomenological nerve model is extended for the neural response to pulsatile
electrical stimulation, which is widely used in cochlear implants as it reduces
inter channel interference.
Neurophysiological data from adult guinea pigs were fitted to the LIFDT model.
First, the parameters were constrained by the Input/output (I/O) curve analysis.
Analysis of the data showed strong accommodation effects. The figures of I/O
function for each pulse were plotted according to the physiological data. Fitting
the I/O function of the data constrained the value of four variables of LIFDT
model. The other five parameters were “optimised by eye”. Although the LIFDT
is built with stimulus-dependent threshold, the response of short duration
biphasic pulsatile stimuli exhibits weak accommodation effects.
Then, in order to avoid the complication of full optimization, analytical
approximation of the LIFDT model was derived for pulsatile electrical
stimulation. It improves computational efficiency and provides information on
how the parameters of the LIFDT model affect the accommodation effects.
Theoretical predictions indicate that the LIFDT model could not capture the
strong accommodation effects in the neurophysiological data due to structural
problems. Alternatively, a Markov renewal process model was utilized to track the pulsetrain
response. The stationary and non-stationary Markov renewal process
models were fitted to the neurophysiological data. Both models can interpret the
conventional PST histograms into conditional probabilities, which are directly
related to the interpulse intervals. The consistent results from those two models
provide a qualitative analysis of the accommodation characteristics.

Item Type:	Thesis (PhD)
Subjects:	Q Science > QP Physiology T Technology > TK Electrical engineering. Electronics Nuclear engineering
Library of Congress Subject Headings (LCSH):	Electric stimulation -- Mathematical models, Acoustic nerve, Neurophysiology -- Data processing, Cochlear implants
Official Date:	April 2012
Dates:	Date Event April 2012 Submitted
Institution:	University of Warwick
Theses Department:	School of Engineering
Thesis Type:	PhD
Publication Status:	Unpublished
Supervisor(s)/Advisor:	Stocks, Nigel G.
Extent:	xix, 207 leaves : charts.
Language:	eng

Request changes or add full text files to a record

Repository staff actions (login required)

View Item

Downloads