David Sengeh Thesis Defense

December 14, 2015


MIT Media Lab, E14-633


For people living with limb amputation, the prosthetic socket – the interface between the residuum and prosthesis – is the most critical component. Where a socket is uncomfortable, especially due to poor fit, the quality of life for a patient is greatly hindered. Yet, conventional design of sockets is largely artisan with limited input of quantitative data. Current computer-aided and manufacturing (CAD/CAM) designs are still not clinically applicable solutions. Due to model identification procedures that employ non patient-specific and incomplete data sets, today’s FEA models of the residuum are not predictive leading to suboptimal socket designs. As such, there exists a need for a comprehensive biomechanical model of the residuum for the quantitative design, and computational evaluation, of patient-specific prosthetic sockets. This thesis presents a combined experimental-numerical approach to evaluate and validate a transtibial residuum biomechanical model. The central hypothesis of the work is that a single biomechanical model can predict the large nonlinear response at various sites on a residuum under load. To evaluate this hypothesis, a nonlinear, two-tissue model was formulated where tissue geometries were defined using MRI data of the residuum. The non-linear viscoelastic mechanical parameters of the model were identified using inverse FEA-based optimization using in-vivo indentation experimental data. Using optimized model tissue parameters, the mean percentage error (mean absolute error/ maximum experimental force) between the experimental and simulation force-time curves at 14 locations across an evaluated transtibial residuum was 7 +/-3%. Using this same modeling methodology and a single set of material constants to describe the bulk soft tissue biomechanical response of seven distinct transtibial residual limb models, the average percentage error for indentations at multiple locations across all seven limbs was 7 +/-1%. From this predictive model of a residuum, a novel multi-material transtibial socket was designed, fabricated and evaluated through an entirely automated and repeatable methodology. In a preliminary clinical investigation, the multi-material socket was shown to reduce the peak contact pressures at critical locations on the residuum during standing compared to a conventional socket interface designed and fabricated by a trained prosthetist.

Host/Chair: Hugh Herr


Neri OxmanNicholas NegroponteKevin Moerman

More Events