| Amputee gait adaptation to mechanical changes in prosthetic componentry is an inferred, but not well understood phenomenon that may be similar to an adaptive strategy used for able-bodied individuals. A two-element spring model, successfully used to predict the efficient excursion of the center-of-mass (COM) and adaptation to surface stiffness in able-bodied subjects was modified to evaluate overall, prosthetic foot, and residual limb compliance of amputees in bouncing gait. Two spring model variants, incorporating either linear or bilinear elements, were used to correlate residual limb compliance of transtibial amputees to the mechanical stiffness of energy storage and return (ESAR) prosthetic feet. A clinical trial of bouncing gait activities was conducted to evaluate overall limb stiffness in three male transtibial amputee subjects using three stiffness categories of ESAR prosthetic feet. Mechanical testing of the feet was conducted to replicate the mechanical behavior of the feet during the clinical trial. The model variants were then used to assess the subjects' residual limb compliance in response to changes in prosthetic foot compliance. Results from the clinical trial and mechanical testing showed (1) subject-dependent bouncing gait styles for various activities, (2) limb stiffness behavior similar to able-bodied subjects hopping below their preferred frequencies, (3) constant overall limb stiffness for two subjects in eleven of twelve comparisons of prosthetic feet, and (4) distinct populations of prosthetic foot stiffness that matched the manufacturer's categories, irrespective of alignments for all activities and subjects. The developed analytical model of amputee gait adaptation showed that: (1) the bilinear variant failed to meet any of the criteria necessary to demonstrate amputee gait adaptation for any subject or activity and (2) the linear variant matched the hypothesized negative correlation between residual and prosthetic limb compliance for one subject and activity. Future research should isolate the phases of ground contact instead of modeling the entire period with a single model. |
