Frameworks for the design of passive prosthetic knee components using user-centered methods and biomechanics of level-ground walking

Abstract

Passive knee prostheses in developing countries use low-cost components driven primarily by the need to prevent falls, resulting in undesirable gait deviations during walking. There is a severe lack of reliable data on the specific needs of low-income amputees, which poses a significant challenge towards developing globally appropriate prosthetic technology. This thesis presents the analysis of user-centered needs and relevant lower leg dynamics as frameworks for the design of passive prosthetic knee components that can enable transfemoral (above-knee) amputees to ambulate with minimal gait deviations leading to higher user satisfaction. The goal of developing these frameworks is ultimately to design a low cost, fully passive prosthetic knee device for persons with transfemoral amputations living in the developing world. To identify user needs, structured oral interviews of 19 transfemoral amputees in India were conducted regarding 22 different Activities of Daily Living (ADLs).

A scale of relative importance for different needs was compiled, which can help designers, doctors, and administrators provide better clinical solutions to amputees. Cross-legged sitting was identified as the most critical user need with the potential for highest improvement in the quality of life of amputees. Two identical rotator prototypes were designed and validated for cross-legged sitting on 9 amputees in India. To compute and replicate the target knee moment profile for a prosthetic knee device, the dynamics of level-ground walking were analyzed using a conceptual link-segment model of the prosthetic leg with the knee joint modeled as a combination of passive linear springs and dampers. The effects of changes in inertial properties (mass, radius of gyration, and center of mass location) of the prosthetic leg on the lower leg kinetics were also quantified in the model.

The knee moment required for achieving normative joint kinematics at the hip, knee and ankle by the optimal engagement of spring and dampers was replicated computationally with a maximum R²=0.90 in an idealized clutching scheme. Multiple prototypes of modular knee mechanisms were built to replicate the model, including (i) an automatic locking module for stability during early stance, (ii) a linear spring module for facilitating knee flexion-extension during early stance, and (iii) a rotary damping module for control during terminal stance and swing. Qualitative feedback from two unilateral transfemoral amputees in India showed the automatic locking module provided the predicted performance for timely stance to swing transition. Fluid-based viscous damping was found to provide more optimal control compared to friction-based damping. A comprehensive biomechanical framework was developed that predicted the range of optimal damping coefficients for transfemoral amputees.

The framework used the results from the link-segment model and empirical data of transfemoral gait characteristics such as slower walking speeds and asymmetries in the stance-swing duration. An experimental prosthetic knee with five different damping conditions was built and tested on three subjects with unilateral transfemoral amputation in a motion capture lab. Increased damping led to reduced peak knee flexion during terminal stance and swing, as predicted by the framework. The framework predicted the optimal damping value for achieving normative peak knee flexion to within one standard deviation of the able-bodied value during the swing phase.