A human analog for benchmarking vestibular tests in a roto-tilt chair

Abstract

This thesis gives a background into the anatomy and physiology of the peripheral sensory organs of the vestibular system, as well as some diagnostic tests of the inner ear. Using previous work in vestibular modeling and prosthetics development, this thesis then brings focus to the development and testing of an experimental model that will be used to simulate the vestibulo-ocular reflex (VOR) about pitch, roll, and yaw axes. A sensor array consisting of accelerometers and gyroscopes was built to simulate vestibular function with the sensors positioned in a manner to replicate anatomic placement of the semicircular canals and otolith organs. The accelerometers and gyroscopes are commercially available micro-electro-mechanical systems (MEMS) devices with ranges of ±18g and ±1500^0 /s, respectively. The sensor array is known as a hardware implementation of a vestibular system (HIVeS). Using a mathematical model of vestibular function developed by Chun and Robinson (1978), eye movements were then simulated using a National Instruments (NI) CompactRIO system driven with NI LabVIEW software. The software simulation is known as the software implementation of the VOR (SIVOR). To test the system, the HIVeS unit was mounted on a roto-tilt chair and stimulated with various maneuvers. The SIVOR software simultaneously predicted the corresponding eye movements that were then compared to results from human subject testing. It was found during testing that the SIVOR was able to reproduce some elements, but not all of the human VOR. The failing is not in the HIVeS unit or in the implementation of the SIVOR software on the CompactRIO. Rather, the limitation is in the actual mathematical model used to simulate the human VOR. In spite of that, the HIVeS/SIVOR pair together provide a valuable tool that will enable validation of different mathematical models in future works. Suggestions for improvement of the physical and mathematical models are made, along with ideas for future work based upon the technology developed during the course of this work. Future ideas include use of accelerometer data to simulate stimulation of the human linear acceleration sensors, the saccule and utricle, the development of models of vestibular deficiencies, and the creation of dummies to be used for educational purposes.