We hypothesize that variability in knee subchondral bone surface geometry will differentiate between patients at risk and those not at risk for developing osteoarthritis (OA) and suggest that statistical shape modeling (SSM) methods form the basis for developing a diagnostic tool for predicting the onset of OA. Using a subset of clinical knee MRI data from the osteoarthritis initiative (OAI), the objectives of this study were to (1) utilize SSM to compactly and efficiently describe variability in knee subchondral bone surface geometry and (2) determine the efficacy of SSM and rigid body transformations to distinguish between patients who are not expected to develop osteoarthritis (i.e. Control group) and those with clinical risk factors for OA (i.e. Incidence group). Quantitative differences in femur and tibia surface geometry were demonstrated between groups, although differences in knee joint alignment measures were not statistically significant, suggesting that variability in individual bone geometry may play a greater role in determining joint space geometry and mechanics. SSM provides a means of explicitly describing complete articular surface geometry and allows the complex spatial variation in joint surface geometry and joint congruence between healthy subjects and those with clinical risk of developing or existing signs of OA to be statistically demonstrated.

No abstract

Cervical spinal injuries are a significant concern in all trauma injuries. Recent military conflicts have demonstrated the substantial risk of spinal injury for the modern warfighter. Finite element models used to investigate injury mechanisms often fail to examine the effects of variation in geometry or material properties on mechanical behavior. The goals of this study were to model geometric variation for a set of cervical spines, to extend this model to a parametric finite element model, and, as a first step, to validate the parametric model against experimental data for low-loading conditions. Individual finite element models were created using cervical spine (C3–T1) computed tomography data for five male cadavers. Statistical shape modeling (SSM) was used to generate a parametric finite element model incorporating variability of spine geometry, and soft-tissue material property variation was also included. The probabilistic loading response of the parametric model was determined under flexion-extension, axial rotation, and lateral bending and validated by comparison to experimental data. Based on qualitative and quantitative comparison of the experimental loading response and model simulations, we suggest that the model performs adequately under relatively low-level loading conditions in multiple loading directions. In conclusion, SSM methods coupled with finite element analyses within a probabilistic framework, along with the ability to statistically validate the overall model performance, provide innovative and important steps toward describing the differences in vertebral morphology, spinal curvature, and variation in material properties. We suggest that these methods, with additional investigation and validation under injurious loading conditions, will lead to understanding and mitigating the risks of injury in the spine and other musculoskeletal structures.

Nucleus replacement technologies are a minimally invasive alternative to spinal fusion and total disc replacement that have the potential to reduce pain and restore motion for patients with degenerative disc disease. Finite element modeling can be used to determine the biomechanics associated with nucleus replacement technologies. The current study focuses on a new nucleus replacement device designed as a conforming silicone implant with an internal void. A validated finite element model of the human lumbar L3–L4 motion segment was developed and used to investigate the influence of the nucleus replacement device on spine biomechanics. In addition, the effect of device design changes on biomechanics was determined. A 3D, L3–L4 finite element model was constructed from medical imaging data. Models were created with the normal intact nucleus, the nucleus replacement device, and a solid silicone implant. Probabilistic analysis was performed on the normal model to provide quantitative validation metrics. Sensitivity analysis was performed on the silicone Shore A durometer of the device. Models were loaded under axial compression followed by flexion/extension, lateral bending, or axial rotation. Compressive displacement, endplate stresses, reaction moment, and annulus stresses were determined and compared between the different models. The novel nucleus replacement device resulted in similar compressive displacement, endplate stress, and annulus stress and slightly higher reaction moment compared with the normal nucleus. The solid implant resulted in decreased displacement, increased endplate stress, decreased annulus stress, and decreased reaction moment compared with the novel device. With increasing silicone durometer, compressive displacement decreased, endplate stress increased, reaction moment increased, and annulus stress decreased. Finite element analysis was used to show that the novel nucleus replacement device results in similar biomechanics compared with the normal intact nucleus.

The objective of this investigation was to develop probabilistic finite element (FE) models of the anterior longitudinal ligament (ALL) and posterior longitudinal ligament (PLL) of the cervical spine that incorporate the natural variability of biological specimens. In addition to the model development, a rigorous validation methodology was developed to quantify model performance. Experimental data for the geometry and dynamic properties of the ALL and PLL were used to create probabilistic FE models capable of predicting not only the mean dynamic relaxation response but also the observed experimental variation of that response. The probabilistic FE model uses a quasilinear viscoelastic material constitutive model to capture the time-dependent behaviour of the ligaments. The probabilistic analysis approach yields a statistical distribution for the model-predicted response at each time point rather than a single deterministic quantity (e.g. ligament force) and that response can be statistically compared to experimental data for validation. A quantitative metric that compares the cumulative distribution functions of the experimental data and model response is computed for both the ALL and PLL throughout the time histories and is used to quantify model performance.