Implant Stem Design Optimization, including Statistical Shape and Density Modelling

Background: Stress shielding of bone around the stem components of orthopedic implants, including total shoulder replacement (TSR) implants, can result in bone resorption, leading to implant loosening and failure. Use of more flexible stems could reduce this effect; however, creating such stems is difficult using traditional biocompatible alloys and conventional manufacturing techniques. Recent advances in additive manufacturing (AM) has enabled the production of parts with complex geometries and stiffness distributions from biomedical metal alloys. Hollow porous TSR implant stems fabricated using metal AM can potentially reduce stress shielding. How to design such implants, however, remains unknown. Rigorous and systematic design optimization techniques have seen little usage in the field of orthopaedics, but are an ideal approach to designing such implant components.

Objectives: to optimize the design of hollow porous metal AM stem components of TSR implants to reduce the effect of stress shielding. This will be achieved through completion of three sub-objectives, described below.

Proposed Methods:
    1. Important geometric features describing the general shape of a stem will be defined in a parametric computer-aided design (CAD) model. Pore boundaries and wall thickness over the entirety of the stem will also be defined and parametrized, and with pore positions and shape defined by a Voronoi distribution. A 3D solid tetrahedral mesh will be developed from the porous surface model, which can be imported into finite element (FE) modeling software.
    2. After implantation of the stem into a proximal humerus bone, a FE model will be developed to predict bone and stem stresses. Following computed tomographic imaging of a cadaveric shoulder, a 3D model of the humerus bone will be segmented and imported into an FE software. Boolean operations will be applied to crop and virtually ream the bone. A 3D tetrahedral FE mesh will be created on the remaining geometry and also the intact bone. The 3D solid mesh of the hollow porous stem will be assembled with the reamed bone. Regional bone material properties will be applied based on CT attenuation. Forces representing typical shoulder loading will be applied to both intact and implanted models and elemental stresses and strain energy densities will be compared. A decrease in either compared to the intact bone can imply stress shielding. Consequently, for at least three different designs, changes in cortical surface strains will be measured at pre-defined locations to be compared with strain gauge measurements described in (3).
    3. Prototypes of the three stem designs chosen for more analysis in (2) will be additively manufactured. The cadaveric humerus which was used to develop the FE models will be mounted to a 6DOF joint motion simulator in order to apply the simulated loads in (2). Strain gauges will be placed at the same locations as in (2). Model-predicted and experimentally-measured strains of the intact humerus will be compared. Subsequently, each of the prototype stems will be implanted and loaded. Measured surface strains will be compared with strains of intact humerus and model predictions.

Significance: The proposed research is expected to develop a validated FE modeling technique for predicting stress shielding in stem and bone composite structures, and also investigate the effects of various porous hollow stem design parameters on stress shielding.

List of submitted abstracts:
  1. Soltanmohammadi, P. and Willing, R. (2019). Development of a Statistical Shape and Density Model of the Shoulder. In the 2nd International Combined Meeting of Orthopaedic Research Societies (ICORS), Montréal, Québec, Canada.
  2. Soltanmohammadi, P. and Willing, R. (2019). Development of a Statistical Density Model of the Shoulder. In the 2019 Annual Meeting of Orthopaedic Research Society (ORS), Austin, TX, USA.

List of papers in progress:
  1. Soltanmohammadi, P., Veeraraghavan, V., Jacobs, D. and Willing, R (2019). Development of a Statistical Shape and Density Model of the Shoulder.