Date of Award
Spring 2024
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Biomedical Engineering
First Advisor
Brian Stemper
Abstract
A wearable ear-based system, incorporating hearing protection, communications capabilities, and a head impact sensor, has been developed for military use. Enhancing computational efficiency is a pivotal goal for system sustainability. The first and second aims of this research were to assess tradeoffs in accuracy for two potential on-board computational processes: 1) rigid body transformations and 2) adjusting transformation dimensions for different head sizes. The third aim of this research study was to validate the performance of the system in the laboratory. For the first aim, ear-measured and transformed ear-to-head CG accelerations were compared to head CG-measured accelerations. For the second aim, ear accelerations of a small and a large individual were modeled and then transformed to their approximate head CG locations via two transformation methods: 1) transformation using standard 50th percentile male head surface-to-head CG dimensions and 2) transformation using subject specific head surface-to-head CG dimensions. Transformed ear-to-head CG accelerations were compared between standard and subject-specific transformation methods. For the third aim, recorded accelerations and accelerations transformed to the head CG of the ear-based system were compared to those of a reference sensor. Transformed ear-to-head CG accelerations were not significantly different than head CG-measured kinematics (p = 0.184), but ear-measured accelerations were (p < 0.05). For both small and large head sizes, transformed ear-to-head CG accelerations based on 50th percentile and actual transformation distances were not significantly different, respectively (p = 0.495, p = 0.409). Recorded accelerations and accelerations transformed to the head CG of the ear-based system were significantly different than those of the reference sensor, respectively (p < 0.05, p < 0.05). These findings indicated that while transformed ear-to-head CG accelerations represented head accelerations and ear-measured accelerations did not, it is not necessary to adjust transformation dimensions for different head sizes. Instead, using standard transformation dimensions for transformed ear-to-head CG accelerations of different head sizes was comparable to those using subject-specific transformation dimensions. Additionally, it was found that the ear-based system did not provide suitable head accelerations due to inadequate coupling to the head.