Pure Tone audiometry is performed routinely in hospital hearing clinics throughout the world. The test machines in general use are manually operated and the tester controls the frequency and intensity of the stimulus. Although the advent of computers has brought a certain amount of computer control into the field, the computers have been used for storing data which has been accumulated using manual machines, or for performing completely automated screening tests which allow the operator no intervention. The work of this thesis tries to bridge this gap by using an automated audiometer which can be interrupted and overruled by the operator at any stage but which still affords all the benefits of the computer in data storage and management. It also investigates and uses aspects of computer control which allow tests to be carried out which would otherwise be too complicated or time-consuming. Chapter 1 offers an introduction to audiology, both from a medical and a scientific point of view and describes some of the work which has been done previously in this field. Chapter 2 explains the implementation under computer control of the conventional pure-tone threshold tests (recommended by the British Society of Audiology). This includes an entirely automated test which can be paused, stopped, restarted or overruled by the operator, if and when required. The system also incorporates the ability to perform a completely manually controlled test. Chapter 3 describes a 'random test' which has been designed and implemented, and which is unique to this machine. It takes a block of points around the patient's thresholds and randomises the presentation of tones in ear, in frequency and in intensity. This random test is then investigated to determine whether it can be used to distinguish subjects who are suspected of exaggerating their hearing loss in order to qualify for industrial hearing loss benefit, from the subjects who are being truthful about their hearing thresholds. A number of normal hearing subjects were asked to simulate a hearing loss and then asked to respond truthfully. Several quantities were measured and the two sets of results are compared. Using various types of analysis, it was found possible to distinguish the truthful set of results from the non-organic hearing loss set. Chapter 4 describes the implementation of computer control to several supra-threshold tests, namely the loudness balance tests, and the tone decay tests. There are many benefits of using a computer to control these tests. Chapter 5 and chapter 6 deal with a different form of audiometry which uses speech as the stimulus instead of pure-tones. Chapter 5 describes the implementation of a conventional speech test. This has the advantage that the words spoken can be displayed on the PC screen by using a coding system for the second channel of a stereo audio tape. Chapter 6 is a modification of this test which, using a mechanism of scoring the phonemes individually, allows a chart of phoneme versus hearing level to be displayed, giving speech processing information. Chapter 7 describes the design and fabrication of a multi - frequency tympanometer. A brief conclusion summarising the work achieved is contained in chapter 8. It describes the benefits afforded by computer control and in particular deals with those tests which would be virtually impossible to undertake manually. Appendix A gives a brief outline of clinical decision analysis, one of the forms of analysis used in chapter 3. Appendix B gives a complete list of the Boothroyd wordlists along with their phonetic components. Appendix C gives a test program written in 'C' which fits a cubic spline curve to the points on a speech audiogram. Appendix D gives the results of a frequency analysis done of each individual phoneme of the words in the Boothroyd wordlists. Appendix E shows two MC68000 assembly code programs used to control the tympanometer. This thesis shows several uses of computers in audiology which until now have remained untapped.
PDF
download
878987797
028-85220240
