Development of Subject Specific Finite Element Models of the Mouse Knee Joint for Preclinical Applications.

Osteoarthritis is the most common musculoskeletal disabling disease worldwide. Preclinical studies on mice are commonly performed to test new interventions. Finite element (FE) models can be used to study joint mechanics, but usually simplified geometries are used. The aim of this project was to create a realistic subject specific FE model of the mouse knee joint for the assessment of joint mechanical properties. Four different FE models of a C57Bl/6 female mouse knee joint were created based on micro-computed tomography images of specimens stained with phosphotungstic acid in order to include different features: individual cartilage layers with meniscus, individual cartilage layers without meniscus, homogeneous cartilage layers with two different thickness values, and homogeneous cartilage with same thickness for both condyles. They were all analyzed under compressive displacement and the cartilage contact pressure was compared at 0.3 N reaction force. Peak contact pressure in the femur cartilage was 25% lower in the model with subject specific cartilage compared to the simpler model with homogeneous cartilage. A much more homogeneous pressure distribution across the joint was observed in the model with meniscus, with cartilage peak pressure 5-34% lower in the two condyles compared to that with individual cartilage layers. In conclusion, modeling the meniscus and individual cartilage was found to affect the pressure distribution in the mouse knee joint under compressive load and should be included in realistic models for assessing the effect of interventions preclinically.

PTH(1-34) treatment and/or mechanical loading have different osteogenic effects on the trabecular and cortical bone in the ovariectomized C57BL/6 mouse.

In preclinical mouse models, a synergistic anabolic response to PTH(1-34) and tibia loading was shown. Whether combined treatment improves bone properties with oestrogen deficiency, a cardinal feature of osteoporosis, remains unknown. This study quantified the individual and combined longitudinal effects of PTH(1-34) and loading on the bone morphometric and densitometric properties in ovariectomised mice. C57BL/6 mice were ovariectomised at 14-weeks-old and treated either with injections of PTH(1-34); compressive loading of the right tibia; both interventions concurrently; or both interventions on alternating weeks. Right tibiae were microCT-scanned from 14 until 24-weeks-old. Trabecular metaphyseal and cortical midshaft morphometric properties, and bone mineral content (BMC) in 40 different regions of the tibia were measured. Mice treated only with loading showed the highest trabecular bone volume fraction at week 22. Cortical thickness was higher with co-treatment than in the mice treated with PTH alone. In the mid-diaphysis, increases in BMC were significantly higher with loading than PTH. In ovariectomised mice, the osteogenic benefits of co-treatment on the trabecular bone were lower than loading alone. However, combined interventions had increased, albeit regionally-dependent, benefits to cortical bone. Increased benefits were largest in the mid-diaphysis and postero-laterally, regions subjected to higher strains under compressive loads.

Estimation of in vivo inter-vertebral loading during motion using fluoroscopic and magnetic resonance image informed finite element models.

Finite element (FE) models driven by medical image data can be used to estimate subject-specific spinal biomechanics. This study aimed to combine magnetic resonance (MR) imaging and quantitative fluoroscopy (QF) in subject-specific FE models of upright standing, flexion and extension. Supine MR images of the lumbar spine were acquired from healthy participants using a 0.5 T MR scanner. Nine 3D quasi-static linear FE models of L3 to L5 were created with an elastic nucleus and orthotropic annulus. QF data was acquired from the same participants who performed trunk flexion to 60° and trunk extension to 20°. The displacements and rotations of the vertebrae were calculated and applied to the FE model. Stresses were averaged across the nucleus region and transformed to the disc co-ordinate system (S1 = mediolateral, S2 = anteroposterior, S3 = axial). In upright standing S3 was predicted to be -0.7 ±â€¯0.6 MPa (L3L4) and -0.6 ±â€¯0.5 MPa (L4L5). S3 increased to -2.0 ±â€¯1.3 MPa (L3L4) and -1.2 ±â€¯0.6 MPa (L4L5) in full flexion and to -1.1 ±â€¯0.8 MPa (L3L4) and -0.7 ±â€¯0.5 MPa (L4L5) in full extension. S1 and S2 followed similar patterns; shear was small apart from S23. Disc stresses correlated to disc orientation and wedging. The results demonstrate that MR and QF data can be combined in a participant-specific FE model to investigate spinal biomechanics in vivo and that predicted stresses are within ranges reported in the literature.

Image driven subject-specific finite element models of spinal biomechanics.

Finite element (FE) modelling is an established technique for investigating spinal biomechanics. Using image data to produce FE models with subject-specific geometry and displacement boundary conditions may help extend their use to the assessment spinal loading in individuals. Lumbar spine magnetic resonance images from nine participants in the supine, standing and sitting postures were obtained and 2D poroelastic FE models of the lumbar spine were created from the supine data. The rigid body translation and rotation of the vertebral bodies as the participant moved to standing or sitting were applied to the model. The resulting pore pressure in the centre of the L4/L5 disc was determined and the sensitivity to the material properties and vertebral body displacements was assessed. Although the limitations of using a 2D model mean the predicted pore pressures are unlikely to be accurate, the results showed that subject-specific variation in geometry and motion during postural change leads to variation in pore pressure. The model was sensitive to the Young׳s modulus of the annulus matrix, the permeability of the nucleus, and the vertical translation of the vertebrae. This study demonstrates the feasibility of using image data to drive subject-specific lumbar spine FE models and indicates where further development is required to provide a method for assessing spinal biomechanics in a wide range of individuals.