Publication Date

12-2021

Date of Final Oral Examination (Defense)

9-27-2021

Type of Culminating Activity

Thesis

Degree Title

Master of Science in Mechanical Engineering

Department

Mechanical and Biomechanical Engineering

Major Advisor

Clare Fitzpatrick, Ph.D.

Advisor

Mahmood Mamivand, Ph.D.

Advisor

Gunes Uzer, Ph.D.

Abstract

In vivo clinical studies are the optimal way to investigate the biomechanical outcomes of new prosthetic devices. This particular style of testing can be difficult and, in certain cases, unethical to perform. The testing of unproven devices, surgical techniques, and materials put patients at risk from unanticipated outcomes in how these devices respond to the in vivo environment and patient-specific loading conditions. Biomechanical computational models were developed to provide validation to new devices prior to clinical testing. Computational models for use in optimizing knee prosthetics frequently include ligament representations, but these representations have inherent uncertainty due to wide intersubject variation across the population and difficulties in defining the subject-specific properties of the ligament. Typically, these parameters are tuned using a trial-and-error approach by experienced personnel or by adapting literature values. However, published values for ligament constants have been shown to vary greatly. Due to these issues, previous studies have recommended the use of a sensitivity analysis to validate input parameters and incorporate uncertainty into finite element (FE) analyses. FE simulations have been widely used in implant development, but this has largely been performed using subjects with a normal body mass index (BMI). Increasing BMI has been shown to negatively impact the outcomes of total knee arthroplasty (TKA) procedures. Better understanding of obesity’s impact on joint kinematics will help with design refinement for prosthetics used in obese individuals.

The first objective of this study was to use dynamic, in vivo kinematics to bound subject-specific ligament parameters to a region that will produce physiological forces, thus removing some uncertainty associated with ligament properties. A Monte Carlo simulation was used to find the region of physiological properties for each subject across multiple dynamic activities. The resulting, subject-specific regions were then compared between cohorts of high BMI and normal BMI subjects. Significant differences were found between the bounding areas for these groups. Additionally, there were ligaments that had significant differences when age and gender were considered. This study indicates that there are likely cohort-specific differences in in vivo ligament properties.

The secondary objective was to create a model database of three high BMI subjects as they perform five activities of daily living. Subject-specific, FE simulations were created for the three subjects using kinematics that were derived from experimental data of the subjects completing various exercises. Controllers were developed and tuned to apply the muscle forces, matching profiles generated from the kinematic data. The models correctly reflect the target profiles and are a crucial first step to analyze the outcomes of prosthetic devices in the obese population. Industry partners are currently using this model database to virtually implant prosthetic devices and measure how the resulting kinematics match the natural, non-implanted knee.

DOI

https://doi.org/10.18122/td/1886/boisestate

Share

COinS