A preliminary comparison of prosthetic socket liner strain determined using digital image correlation and finite element analysis

by Mohammadreza Freidouny, Michael L. Madigan, Carson Squibb, Masaki Hada, Abbie Bailey, Trevor Johnson, Brian Kaluf, Michael K. Philen

Improving prosthetic socket fit remains an important issue among prosthesis users. Understanding strains at limb-socket interface may help prosthetists improve socket fit. Previous studies have used finite element analysis (FEA) to predict stresses and strains at limb-socket interface. However, the selection of FEA model material properties as well as the validation of FEA strain results remain challenging. Digital image correlation (DIC) is a well-accepted optical technique that is widely used in experimental mechanics to noninvasively measure strains and can be used to validate FEA predictions. The goal of this feasibility study was to demonstrate the use of DIC to measure prosthesis liner strains of a unilateral transtibial prosthesis user and conduct a preliminary comparison of these strains with those estimated from an FEA model. One participant with a transtibial residual limb was recruited. Liner strains of the residual limb during two stepping tasks were measured with DIC. An FEA model of the residual limb was also developed, and its estimates of liner strains were compared with DIC measurements. Results showed that mean and probability distributions of principal strain values measured by DIC were in moderate to good agreement with FEA estimations. For example, DIC measured mean maximum and minimum principal strains of 0.041 and 0.021 in the anterior aspect of the residual limb when it was being vertically loaded, and FEA estimated mean maximum and minimum principal strains of 0.051 and 0.027 in the same region. The Bhattacharyya coefficient (a measure of overlap between two probability distributions, with values approaching 1 indicating high similarity) exceeded 0.90 for principal strain distributions between DIC and FEA on the anterior aspect of the residual limb during both tasks. These findings demonstrate that DIC can be used to refine FEA model material properties and results. As such, this methodology may help to advance the design of prosthetic sockets; however, as this study involved a single participant, findings are preliminary and intended to demonstrate methodological feasibility rather than generalize across the broader population of prosthesis users.