Abstract
Microprocessor controlled knee prostheses require expertise and programming skill to setup the swing phase damping performance for different walking speeds. There is a lack of suitable sensing and analysis means to quantify control performance and assist the setup process thereby hindering the development of self-optimizing prostheses. In this project the application of miniature gyroscopes was explored as a means to quantify the swing phase control of transfemoral (TF) amputees. The stride-to-stride repeatability of locomotion and the timing and coordination of knee segment motions were investigated as potential indicators of prosthetic swing control performance. An assessment of the stride to stride repeatability of TF amputee locomotion using the IP+ swing phase control showed that levels of gait repeatability can vary considerably between amputees, are sensitive to changing walking speed and can depend on the run data considered. Comparative analysis of locomotion repeatability and user perception of control was conducted for both conventional and speed adaptive swing phase controls. The speed adaptive control system was found to produce the more repeatable gait at faster and slower than normal walking speeds and was preferred by all the amputees tested. The results indicate that analysis based on the repeatability of locomotion is useful to evaluate swing phase damping setup and control performance. In order to more specifically examine the effects of damping setup on knee control, the timing of knee segment motions were studied. The results revealed a distinctive pattern of waveform time shifting and a distinctive voluntary control mechanism which is believed to be used to influence the timely recovery of the swinging leg. The variability of the timing shifts observed relative to changes in damping control, suggest that a waveform timing analysis method of analysis on it’s own is unlikely to be sufficient for the development of self-optimizing prostheses. Overall the study outcomes support the idea that for each walking speed there is a specific optimum knee motion pattern preferred by amputees and influenced by damping setup. This study highlights the potential of using gyroscopes and segment motion analysis for the study of the control of TF amputee gait. Based on these findings the use of gyroscopes to develop a new generation of knee control systems that actively self-optimizes user voluntary control by quantifying changes to voluntary control directly is proposed for future work.