Maneuvering your limbs requires both accurate commands for how to move, and accurate feedback of their true movements. Conventional prosthetic arms currently lack this sense of proprioceptive feedback, which can make daily tasks difficult without close visual monitoring. Although studies have successfully provided artificial proprioceptive feedback to improve control, this benefit is only consistent when devices are out of view. When sight of the prosthesis is unobstructed, improvements are inconsistent across studies. A potential explanation is that vision may provide more precise proprioceptive information than artificial feedback; thus, artificial feedback may not significantly improve one’s understanding of their prosthesis movements if visual feedback is present. In this thesis, I studied human vision to better understand how well it estimates different aspects of biomimetic movements, such as endpoint and joint speeds in stationary and moving reference frames. Through a series of psychophysics experiments, I found that distal joint speed measured relative to proximal joints (e.g. elbow speed in relation to a moving shoulder) is perceived with high uncertainty. I then developed a sensory substitution system using audio feedback to supplement vision and reduce this joint speed uncertainty. In virtual center-out reaching tasks emulating a myoelectric prosthesis, I demonstrated that non-amputee subjects achieved lower baseline errors when the dynamics of the arm were perturbed if provided joint speed feedback. These results were more pronounced for transradial amputee subjects, who demonstrated lower reaching errors with audio feedback in the presence and absence of perturbations. These results suggest that the availability and precision of intact senses is a crucial factor when developing sensory feedback for prosthetic arms. When intact senses are considered alongside other aspects of the task, we may be able to develop reliable sensory feedback systems for prosthesis users to improve proprioception and the ability to complete everyday tasks.

Creator

• Earley, Eric Joseph

DOI

• https://doi.org/10.21985/n2-ed2r-2a20

Subject

• Biomedical engineering
• Neurosciences

Language

• en

Alternate Identifier

• etdadmin_upload_713001
• http://dissertations.umi.com/northwestern:14963

Keyword

• motor adaptation
• sensory feedback
• myoelectric prosthesis

Date created

• 2019-01-01

Resource type

• Dissertation

Rights statement

• In Copyright