Abstract
Although there have been considerable developments in surgical methods and prosthetic limbs, one quarter of transfemoral (TF) amputees remain dissatisfied with their prostheses. Function could be improved by establishing a communication link between the prosthetic limb and the residual limb. A potential non-invasive method of linking the limbs could be made by using surface electromyography (sEMG). The recent development of the osseo-integrated prosthesis (OP) allows electrodes to be freely attached without interference from the socket. This study’s objective was to investigate patterns of sEMG in TF amputees fitted with OPs during the stance phase of gait. Two groups of subjects were investigated consisting of five TF amputees fitted with OPs and ten subjects with intact limbs. sEMG was recorded from five hip muscles of the amputee’s residual limb or the intact subject’s left limb as the subject walked along a level force platform. Kinematic data was also recorded from the amputee subjects. Prior to walking, the subjects were asked to perform isometric contraction of the recorded muscles. Four types of sEMG variables were determined for the amputees from the stance data. Principle component analysis (PCA) and cluster analysis were then applied to these variables. All the amputee group’s muscles, both cleaved and intact, displayed increased activity during isometric muscle contraction. The sEMG recorded during the stance phase in the intact group corresponded to that reported in published literature. The sEMG measured from the amputees was similar to that of the intact group with the exception of adductor magnus. Inter-stance subject repeatability was low for both groups. The repeatability of stance sEMG for the intact group was lower than that published in previous studies based on the entire gait cycle. PCA of the sEMG for the amputee group revealed major co-contraction of muscles. Cluster analysis of multi-muscle matrices demonstrated significant cluster patterns for the stance phase. Cluster patterns showed little change when repeat measurements were taken eleven months later. The findings suggest that sEMG can potentially be harnessed to activate a micro-computer controlled prosthetic knee joint during locomotion for certain subjects.