Design and Clinical Evaluation of a Digital Transtibial Prosthetic Interface

Abstract

For those living with lower-limb loss, the prosthetic interface, comprising a socket and liner, is the component of the prosthesis that most limits its wearability and use. An improperly designed prosthetic interface results in excessive pressure areas that cause wear and chafing with skin breakdown as a common occurrence. Traditionally designed interfaces require extensive time from the patient and an experienced prosthetist, with these factors compounding to make the entire process inaccessible to the majority of persons with amputation. To address these problems, this thesis outlines a prosthetic interface design and manufacturing pipeline that uses a novel computational algorithm to create subject-specific transtibial liner and socket components that can be additively manufactured at low cost. The residual limb is imaged using a magnetic resonance imaging (MRI) device, and the image set is segmented into a three-dimensional model. This approach is superior to other 3D-modeling prosthetic interface techniques as it is able to capture bone geometries and soft tissue depths of the residuum. A more accurate topology of the skin is captured using digital image correlation (DIC), and this mesh is used in replacement of the MRI skin. The socket is divided into four distinct pressure regions, and the nominal pressure applied at each region can be adjusted to be patient-specific. Finite element analysis is run to simulate liner donning and bodyweight loading upon the interface to generate the final pressure map and liner-socket geometries. Novel prosthetic interfaces made using this algorithm were evaluated against conventionally made interfaces for 5 limbs from 4 patients through a combination of kinematic gait data, standing pressure data, thermal skin measurement, and qualitative patient response. The kinematic results in this study use the Mahalanobis distance to evaluate difference in gait asymmetry resulting from conventional and novel prosthetic interfaces. The distance is calculated using asymmetries for step time, swing time, and peak impact ground reaction force. No subjects exhibit significant difference in gait asymmetry resulting from conventional and novel prosthetic interfaces (asymmetry greater than the 5% p-value was not observed for Mahalanobis distance for 3 degrees of freedom). Thermal results show no statistically significant difference in percent temperature change from reference between conventional and novel interfaces. This is true for overall temperature change as well as change at the distal and fibular head regions specifically. Further, standing pressure data do not show significant difference between conventional and novel prosthetic interfaces when the pressure variance at locations excluding the patellar bar are compared. Qualitative feedback from the three unilateral subjects participating in the study are generally neutral, with novel interfaces being evaluated as close in fit to conventional interfaces during sitting and standing. One bilateral patient rates the novel interface as better than the conventional interface on both legs. The three unilateral patients give the novel interfaces slightly worse ratings while walking, however often comfort was reduced due to unfamiliarity with the socket suspension system or socket material, neither of which are directly applicable to our design. Overall, study results show that the performance of the novel interface is comparable to that of the conventional interface with the potential of providing benefits in overall design time, repeatability, and cost.