Document Type

Thesis

Date of Degree

Fall 2011

Degree Name

MS (Master of Science)

Degree In

Biomedical Engineering

First Advisor

Donald D. Anderson

Second Advisor

Thomas D. Brown

Abstract

Contact stress exposure is thought to play a significant role in many aspects of joint degradation and pathology. Effective and accurate contact stress computation in native or pathological subject specific joints is an important tool in determining the role of contact stress in OA onset and worsening as well as eventually developing and monitoring interventions to prevent joint degradation. In the past FEA modeling has allowed for studies to be completed which relate contact stress exposure human ankle joint to the presence of radiographic OA. While promising, contact FEA for subject specific models is significantly limited by the number of cases that can be computed due to the difficulty of FEA modeling, as well as numerical convergence issues present in contact FEA. To obtain truly statistically powerful conclusions about the causes of joint degradation and OA onset large numbers of subject specific models will need to be created, run and analyzed. Rigid body spring modeling or RBSM has proven to be an effective method of contact stress measurement for both expedited evaluation of PTOA risk following tibial plafond fractures as well as for evaluation of BMLs worsening in a cohort of 38 at risk patients. RBSM treats cartilage as a bed of springs attached to an underlying rigid bone surface. It is a significant simplification from FEA in that it does not allow computation of internal stresses of an object, elaborate material treatments, or true deformation of an object. This simplification comes with the benefit of reduced computational and investigator burden due to the lack of numerical convergence issues as well as no FEA meshing step. A custom written RBSM algorithm was created in MATLAB which works in conjunction with a load balancing algorithm to iteratively solve contact solutions in both load and displacement control. This algorithm was first validated against a previously done physical validation study using two human cadaver ankles in a custom built fixture. The RBSM method was then used to replicate previously obtained FEA results in a study of 22 human ankle joints following tibial plafond fracture. FEA models and loadings were adapted for the RBSM method and run. The RBSM offered a significant speed increase while maintaining comparable results to the FEA. The ability of RBSM to predict PTOA development using a contact stress-time-area exposure metric was virtually unchanged (95% KL grade concordance and 100% OA concordance vs. 94% KL grade concordance and 100% OA concordance, for RBSM and FEA, respectively). The RBSM method was then combined with a feature based 3D-2D alignment routine custom written in MATLAB. This alignment routine uses a ray casting method to recreate a virtual x-ray silhouette edge for a 3D model. This model is then aligned to a 2D edge tracing based off an input radiograph depicting a functional pose of the bone. A global optimizer (simulated annealing) is used to determine the best Euler transform to place the bone in an accurate position in the recreated virtual scene. 38 subject specific knee models segmented from the MOST cohort were aligned to functional appositions bases off of fixed flexion standing radiographs. Contact stresses were then obtained from these aligned joints using RBSM to evaluate the relationship between contact stress level and bone marrow lesion worsening. It was found that as contact stress level increases so does the risk of BMLs worsening. As the worsening of BMLs is associated with joint pain, degradation, and pathology an expedited contact stress method which can accurately predict BML worsening is especially valuable.

Keywords

Ankle, Arthritis, Contact Stress, Knee, PTOA

Pages

x, 110 pages

Bibliography

Includes bibliographical references (pages 103-110).

Copyright

Copyright 2011 Andrew Kern

Share

COinS