Arthur Petron Thesis Defense

December 16, 2015


MIT Media Lab, Bartos E15-070


The prosthetic socket, the mechanical interface between an amputated residuum and an external prosthesis, is of critical importance to the performance of a prosthetic limb system. Conventional prosthetic socket technology is derived using a non-quantitative, artisan methodology. Consequently, a comfortable socket interface cannot be made reproducibly, and persons with limb amputation too often experience discomfort. As a resolution to this difficulty, the field of digital prosthetic socket design seeks to advance a quantitive CAD/CAM methodology for socket production to produce reproducible and comfortable interfaces. Prosthetic researchers have proposed a digital socket production work flow comprising the steps of 1) assessment of residuum tissue biomechanics; 2) modeling optimization of the residuum-socket interface, and 3) fabrication of a variable-impedance socket system based upon these optimizations. In this thesis, two novel technologies are designed, built and evaluated at either end of this work flow, namely a multi-indenter device for in vivo biomechanical tissue measurement and a Quasi-passive variable-impedance transtibial socket interface.
An active indenter platform called the FitSocket is presented. To assess residual-limb tissue biomechanics, the FitSocket comprised 14 position and force-controllable actuators that circumferentially surround a biological residuum to form an actuator ring. Each actuator is individually controllable in position (97.1μm accuracy) and force (330mN accuracy) at a PC controller feedback rate of 500Hz, allowing for a range of measurement across a residuum. At five distinct anatomical locations across the residual limb, force versus deflection data are presented, demonstrating the accuracy and versatility of the FitSocket for residual- limb tissue characterization. A passive, single indenter version of the FitSocket, called the FitPen, is also presented. The FitPen is designed to be ultra-portable in order to take biomechanical measurements in the field outside the laboratory setting.
A quasi-passive socket (QPS) is presented having spatially and temporally varying socket wall impedances. The QPS is an autonomous computerized transtibial prosthetic interface that can stiffen or become compliant using computer-controlled electrolaminate actuators. The QPS measures forces applied by the limb on the socket, 3-axis acceleration of the socket, and the position of the electrolaminates. On a test participant with transtibial amputation, the socket was evaluated through sit-to-stand tests to determine the viability of computer-controlled electrolaminate engagement, and through a walking study to evaluate the ability of the electrolaminates to maintain their clutched state during ambulation at a self-selected walking speed. The average deflections of forced tibia movement in the sit-to-stand tests were 7 ± 2mm while sitting with the electrolaminates in an unclutched state, and 2.1 ± 0.6 mm while standing with the electrolaminates in a clutched state. Further, the walking study showed a maximum unclutched deflection (3.7 ± 0.9 mm)16 times larger than that of the maximum deflection while clutched.

Host/Chair: Hugh Herr


Neri OxmanJoe Paradiso

More Events