Design and development of a sensory feedback system for transradial amputees using body powered prostheses

Gordon-Grant, Catherine

Design and development of a sensory feedback system for transradial amputees using body powered prostheses

Master Thesis

2021

Abstract

Introduction Although significant advancements have been made in the realm of prosthesis design over the last two decades, amputees still reject their prostheses largely owing to the lack of sensory feedback received. The aim of this study was to develop a non-invasive sensory feedback system, including proprioceptive feedback, to work in conjunction with an existing body powered prosthesis that provides the user with increased tactile and proprioceptive awareness demonstrated in the form of improved results in an object identification and size discrimination (OISD) test, improved percieved confidence and improved response time. Materials & Methods A sensory feedback system was developed using principles of rapid prototyping and testing. The system utilises capacitive sensors to mediate detection of proximity and touch, and makes use of a flex sensor to represent the position and motion of the finger digits of the prosthesis. A vibrotactile armband on the upper arm is used to feedback the sensed information to the user. Four able-bodied, adult volunteers participated in a preliminary study to test the efficacy of the capacitive sensor-vibrating motor disc combination. Five able-bodied, adult volunteers participated in a study conducted at the Medical Devices Lab at the University of Cape Town. The study investigated whether using the designed sensory feedback system in conjunction with a body powered prosthesis improves the participant's ability to locate and discriminate the size of spheres whilst visually and aurally restricted. Results & Observations The results from the preliminary study show that all participants were able to distinguish between 3 levels of touch (proximity, light and hard touch) intensity and locate the position of the stimulation for good electrical conductors such as metal, human skin, wood and sometimes ceramic and glass materials depending on the coating. Results from the object identification test indicate that participants performed worse when no sensory feedback was provided (at an average accuracy of 0%). All participants were able to identify the objects accurately and efficiently when sensory feedback was provided at an average accuracy of 100%. Similarly, average accuracy for object discrimination improved from 58.33% with no sensory feedback to 85% with sensory feedback. Response times for both object identification tasks and size discrimination tasks also improved illustrated by a significantly negative relationship (negative gradient) in all trend lines over the period of testing. This suggests a learnability with the device: as the user becomes more cognisant of the device, the better their performance is in accuracy and response time. Confidence levels for size discrimination were highest with sensory feedback active. There was a significantly positive correlation (positive gradient) between accuracy and confidence. With no sensory feedback, confidence was on average 70% which increased to 84% with sensory feedback. Conclusions The non-invasive sensory feedback system met all design requirements successfully. Future recommendations include minor adjustments to the design to optimise its capabilities after which, the design can be revalidated and tested on transradial amputees. Future work includes performing usability studies with amputee subjects, revalidation of the design through bench testing and with amputee subjects, and evaluating the scalability and manufacturability of the design with a variety of commercially available body powered prostheses.

Keywords

Biomedical Engineering

Reference:

Collections

Masters

Full item page