Patient-specific Biomechanical Model Research Articles

An efficient two-stage approach for image-based FSI analysis of atherosclerotic arteries

Patient-specific biomechanical modeling of atherosclerotic arteries has the potential to aid clinicians in characterizing lesions and determining optimal treatment plans. To attain high levels of accuracy, recent models use medical imaging data to determine plaque component boundaries in three dimensions, and fluid-structure interaction is used to capture mechanical loading of the diseased vessel. As the plaque components and vessel wall are often highly complex in shape, constructing a suitable structured computational mesh is very challenging and can require a great deal of time. Models based on unstructured computational meshes require relatively less time to construct and are capable of accurately representing plaque components in three dimensions. These models unfortunately require additional computational resources and computing time for accurate and meaningful results. A two-stage modeling strategy based on unstructured computational meshes is proposed to achieve a reasonable balance between meshing difficulty and computational resource and time demand. In this method, a coarse-grained simulation of the full arterial domain is used to guide and constrain a fine-scale simulation of a smaller region of interest within the full domain. Results for a patient-specific carotid bifurcation model demonstrate that the two-stage approach can afford a large savings in both time for mesh generation and time and resources needed for computation. The effects of solid and fluid domain truncation were explored, and were shown to minimally affect accuracy of the stress fields predicted with the two-stage approach.

Validation of an adapted MRI pulse sequence for quantification of fatty infiltration in muscle

The long-term goal of our research is to develop a patient-specific biomechanical model of the supraspinatus muscle and tendon for analyzing the effects of fatty infiltration and pennation angle changes. One input of the model will be the amount and distribution of fat within the muscle. Therefore, the objective of this project was to adapt and validate a novel magnetic resonance imaging pulse sequence for quantifying the extent of fatty infiltration for the supraspinatus muscle. Samples of incremental combinations of muscle and fat from a cow were prepared and scanned with our quantitative MR technique. The MR results were compared to actual fat measurements taken using a Soxtec extraction analyzer. The MR fat quantification results closely match those values determined by Soxtec analysis (R(2)=.98, root mean square deviation=4.9). Our MR fat quantification protocol can be used to accurately quantify fatty infiltration in skeletal muscle. This ability to evaluate skeletal muscle tissue noninvasively will improve the efficacy of evaluation of fatty degeneration in assessing tissue quality preoperatively. This can be important in surgical planning of any musculoskeletal repairs, particularly rotator cuff tears.

Heterogeneous meshing and biomechanical modeling of human spine

We aim to develop a patient-specific biomechanical model of human spine for the purpose of surgical training and planning. In this paper, we describe the development of a finite-element model of the spine from the VHD™ Male Data. The finite-element spine model comprises volumetric elements suitable for deformation and other finite-element analysis using ABAQUS. The mesh generation solution accepts segmented radiological slices as input, and outputs three-dimensional (3D) volumetric finite element meshes that are ABAQUS compliant. The proposed mesh generation method first uses a grid plane to divide the contours of the anatomical boundaries and its inclusions into discrete meshes. A grid frame is then built to connect the grid planes between any two adjacent planes using a novel scheme. The meshes produced consist of brick elements in the interior of the contours and with tetrahedral and wedge elements at the boundaries. The nodal points are classified according to their materials and hence, elements can be assigned different properties. The resultant spine model comprises a detailed model of the 7 cervical vertebrae, 12 thoracic vertebrae, 5 lumbar vertebrae, and S1. Each of the vertebrae and intervertebral disc has between 1200 and 6000 elements, and approximately 1200 elements, respectively. The accuracy of the resultant VHD™ finite element spine model was good based on visual comparison of volume-rendered images of the original CT data, and has been used in a computational analysis involving needle insertion and static deformation. We also compared the mesh generated using our method against two automatically generated models; one consists of purely tetrahedral elements and the other hexahedral elements.

The relationship between bone mineral density and biomechanics in patients with osteoporosis and scoliosis

Nearly one-third of all women and one-sixth of all men over age 65 have osteoporosis, and this condition is often accompanied by lumbar scoliosis. Previous work has shown that, in a group of postmenopausal women with scoliosis and osteoporosis, both the bone mineral content (BMC) and bone mineral density (BMD) were greater on the concave side than the convex side. The goal of this study was to examine the structure-function relationships in the spines of patients with low bone mass and scoliosis using a patient-specific biomechanical model. We compared the percent change in BMC and the percent change in BMD with axial force, F(a), shear force, F(s), moment, M, local curvature, theta(rel), and the patient's age, A. We found that the percent change in BMC depended on the applied moment and the local curvature. The same dependence was observed for the percent change in BMD, but in this case, the shear force was also significantly inversely correlated. A population with femoral neck BMD with a T-score greater than -2.0 was similarly evaluated and yielded similar results. The percent change in BMD was related to M, theta(rel), A and negatively to the shear force. These results indicate that the osteoporotic spine is still able to respond to changes in the mechanical environment and provides a useful comparison between patients with osteoporosis and those with normal bone mass. In addition, this model may be a useful tool for the in vivo assessment of bone density changes in response to mechanical stimuli and drug treatments.

Probabilistic design analysis of the influence of material property on the human cervical spine.

Studies reported previously in the literature have described the importance of material variation on the cervical responses and have examined some effects by varying the material properties, but there is no systematic approach using statistical methods to understand the influence of material variation on a cervical spine model under a full range of loading conditions, especially under compression and anterior and posterior shear. A probabilistic design system based on Monte Carlo simulation methods using Latin hypercube sampling techniques is used to analyze the material sensitivity of a C4-C6 cervical spine model involving 13 uncertain input parameters on the biomechanical responses and disc annulus stresses under compression, anterior shear, posterior shear, flexion, extension, lateral bending, and axial rotation. The loading types and range of values were as follows: compression, 0-1 mm; anterior shear, 0-2 mm; posterior shear, 0-3.5 mm; flexion, extension, lateral bending, and axial rotation. 0-1.8 Nm with 73.6-N preload. For each case, the load-deflection and key stress values at various spinal components were captured after each load step. The model was also validated under the same conditions. The minimum and maximum predicted responses were within the range of the experimental data. Ignoring compression loading, the combined effects on the biomechanical responses of the cervical ligaments under the remaining loads are enormous. Their total impacts are almost equal to or slightly less than the influence of disc annulus. Results show that the fiber mechanical properties did not have a significant effect on the compressive stiffness. This study reveals important features that help us identify the critical input parameters and enable us to reduce the development time of a patient-specific biomechanical model.

Détermination d'un modèle biomécanique du cerveau par l'analyse d'images: application à la maladie de Parkinson

Stereotactic neurosurgery for Parkinson's disease consists of stimulating deep nuclei of the brain. Although target coordinates are available with high accuracy on the pre-operative Magnetic Resonance Images (MRI), the leakage of cerebrospinal fluid (CSF) during the procedure may lead to a brain deformation and may cause localization errors with respect to the surgical planning. In this paper, we propose a patient-specific biomechanical model of the brain able to recover its global deformation during this type of neurosurgical procedure. Such a model could be used to update the pre-operative planning and to predict the mechanical effects of the intra-operative brain shift.