Lower extremity prosthesis fit results from a complex biomechanical interaction at the interface between the residuum and socket. The quality of fit determines subject comfort, acceptance, and potential for ambulation. The long term goal of this research is comprehensive assessment of lower limb prosthesis fit that is highly correlated with subjective metrics and has high predictive value for complications and functional outcome. In situ static 3D determination of prosthesis socket fit by a valid, repeatable, practical, and comprehensive method is sought to aid prosthesis design and evaluation, and to improve outcome. Volumetric imaging based on x-ray spiral CT will be used for static in situ lower limb prosthesis evaluation with and without axial loading. From these image volume data sets, tissue composition volume fractions (fat, muscle, bone, skin, other) will be extracted, mass properties estimated, and shape information mapped onto a 3D display. Shape and fit will be visualized by 3D. The measurement methods will be validated in vitro with phantom test objects and cadaver parts, and in vivo with adult amputees. Tissue composition estimates will be tested on extremity remnants phantoms, and on cadaver limbs. Prediction of the biomechanical characteristics for a given socket and limb remnant will be achieved by mathematical modeling. Using the CT volume data, we will synthesize polynomial version (p-version) finite element models that are static, fully 3--dimensional, incorporate separate tissue compartment (bone, muscle/fascia, fat, skin, and prosthesis), anisotropic, and nonlinear with large deformations. The models will be validated experimentally and used to predict the quality of socket fit. Measurements using x-rat spiral CT imaging and finite element tools will be used in a logistic regression model to determine the biomechanical characteristics associated with good and poor fitting prostheses. The feasibility of diagnostic fit evaluation using x-ray spiral CT and finite element mapping methods will be determined through ROC and repeatability analysis. Image-derived fit measures will be tested for correlation with subjective reports corrected for covariation due to age, gender, race, nutritional status, diabetes, smoking, amputation characteristics (stump length, reason for amputation), and other factors known to influence fit. On completion, this project will provide a valid and repeatable practical comprehensive volumetric image-based method to aid lower limb prosthesis design and evaluation. Prediction of the biomechanical characteristics of a given socket and limb remnant by p-version finite element modeling will be developed and tested.
