Date of Award


Document Type

Masters Thesis

Degree Name


Organizational Unit

Daniel Felix Ritchie School of Engineering and Computer Science

First Advisor

Mohammad A. Matin, Ph.D.

Second Advisor

James N. Hagler, Ph.D.

Third Advisor

Roger Salters


ABR, Cochlear implant, Electrical stimulation


In the last thirty years, cochlear implants have become an invaluable instrument in the treatment of severe-to-profound hearing impairment. An important aspect of research in the continued development of cochlear implants is the in vivo assessment of signal processing algorithms intended to improve perception of speech and other auditory signals. In trying to determine how closely cochlear implant recipients process sound relative to the processing done by a normal auditory system, various assessment techniques have been applied. The most common technique has been measurement of auditory evoked potentials (AEPs), which involves the recording of neural responses to auditory stimulation. Depending on the latency of the observed response, the evoked potential indicates neural activity at various ascending neurological structures of the auditory system. Although there have been a number of publications on the topic of AEPs in cochlear implant subjects, there is a need for better measurement and research techniques to obtain more in-depth information to facilitate research on effectiveness of signal processing approaches in cochlear implants.

The research presented herein explored the use of MatLab for the purpose of developing a model for electrically evoked auditory brainstem responses (EABRs). The EABR is commonly measured in hearing-impaired patients who have cochlear implants, via electrical stimulation delivered from electrodes in the implanted array. The simulation model developed in this study took as its input the stimulus current intensity level, and used function vectors and equations derived from measured EABRs, to generate an approximation of the evoked surface potentials. A function vector was used to represent the combined firing of the neurons of the auditory nervous system that are needed to elicit a measurable response. Equations were derived to represent the latency and stimulus amplitude scaling functions. The simulation also accounted for other neural activity that can be present in and contaminate an ABR recording, and reduced it through time-locked averaging of the simulated response.

Predicted waveforms from the MatLab model were compared both to published waveforms from a cochlear implant recipient, and a series of EABR waveforms measured by the author in other cochlear implant recipients. Measurement of the EABRs required specialized interfacing of a commercial recording system with the signal processors of the patients' cochlear implants. A novel measurement technique was also used to obtain more frequency-specific information than usually obtained. Although the nonlinearities normally present in the auditory system were not considered in this MatLab simulation, the model nevertheless performed well and delivered results comparing favorably with the results measured from the research subjects.

Publication Statement

Copyright is held by the author. User is responsible for all copyright compliance.

Rights Holder

Douglas A. Miller


Received from ProQuest

File Format




File Size

111 p.


Electrical engineering, Biomedical engineering