Image-based modeling of nonlinear and anisotropic intervertebral disc mechanics

描述（由申请人提供）：本申请涉及广泛的挑战领域（06）使能技术，以及具体的挑战主题，06- eb -109：模型驱动的生物医学技术开发。肌肉骨骼组织的结构模型计算内应力和应变是研究损伤和失效机制的重要技术。此外，准确的模型为预测手术干预、器械植入和组织工程的后果提供了理解。该建议解决了椎间盘建模的挑战。三十多年来，有限元模型已被证明是研究圆盘力学的重要工具。然而，目前可用的圆盘模型有三个显著的局限性，这将在本提案中解决，以提高预测力学行为的保真度。首先，输入模型的纤维环（AF）特性是基于切除的组织测试。切除组织会破坏纤维结构，这会在材料性能测量中引入伪影，并且不代表椎间盘成分和微结构逐渐改变所反映的不均匀性。其次，切除组织样本消除了子结构之间的界面。第三，当前的圆盘模型使用“电缆”元素，这些元素嵌入AF的各向同性实体元素中。表面上，这代表AF结构；然而，自动对焦的组成和微结构是复杂的，需要一个能更好地反映自动对焦各向异性、非线性和非均匀性的本构公式。本提案的目的是开发一种基于图像的椎间盘模型，该模型结合了组织的非均匀性、各向异性和非线性材料特性。我们将使用复杂的使能技术获得具有空间分布材料属性的3D结构模型。目的1：将已建立的二维MRI扩展到高分辨率的三维MRI。评估脉冲序列的性能，并根据需要调整序列参数，以优化分辨率、对比度和信噪比。目标2：在包括压缩、弯曲和扭转在内的多轴加载配置下对人体椎间盘进行成像，并使用最先进的差分图像配准计算3D应变图。目标3：应用反方法建立本构公式和测量应变图，以确定材料的性能。通过预测未在性能计算中使用的加载配置下的应变来验证模型和材料性能。采用不同的加载方式进行参数回归，重复计算研究。在这项研究完成后，我们将能够准确地预测椎间盘内的应力和应变，这是由于纤维成分和纤维外基质的微观结构贡献，以及椎间盘亚结构之间的相互作用。原位确定材料特性是一种创新的解决方案，解决了目前切除组织测试中人工制品的局限性和亚结构界面的破坏。这项工作的意义在于开发了一种宝贵的工具来精确地模拟椎间盘力学，这对于理解椎间盘的机械故障和治疗它都是至关重要的。该系统将首先应用于人体尸体组织；然而，影像学的长期进步可能允许扩展到体内应用和患者特异性建模和分析。本挑战提案中所描述的拟议实验将在宾夕法尼亚大学（2.5名博士后研究人员，0.2名技术人员）和明尼苏达大学（1名研究生，0.5名博士后研究人员）之间的地点创造4个全职职位。宾夕法尼亚大学医学院在创造新知识和新疗法以改善人类健康以及培养下一代科学领袖方面是国际公认的领导者。在追求这些目标的过程中，宾夕法尼亚大学医学院在宾夕法尼亚、新泽西和特拉华州广泛而深刻地扩展了其经济影响。最近的研究表明，2008年宾夕法尼亚大学医学院创造了37,000个就业岗位和54亿美元的经济活动。作为明尼苏达州唯一的研究型大学，明尼苏达大学是该州的经济引擎。因此，明尼苏达大学的影响是深远而广泛的，有超过21,500个工作岗位与授予明尼苏达大学的研究资助直接相关。本申请涉及广泛的挑战领域（06）使能技术，以及具体的挑战主题，06- eb -109：模型驱动的生物医学技术开发。计算椎间盘内应力和应变的结构模型对于研究椎间盘损伤、撕裂和失效的机制至关重要。此外，准确的模型将为预测手术干预、器械植入和组织工程的后果提供理解。不幸的是，目前可用的圆盘模型缺乏准确预测力学行为的保真度。本提案的目的是开发一种基于图像的椎间盘模型，该模型结合了组织的非均匀性、各向异性和非线性材料特性。本挑战提案中所描述的拟议实验将在宾夕法尼亚大学（2.5名博士后研究人员，0.2名技术人员）和明尼苏达大学（1名研究生，0.5名博士后研究人员）的站点之间创造4个全职职位。

DESCRIPTION (provided by applicant): This application addresses the broad Challenge area (06) Enabling Technologies, and the specific Challenge Topic, 06-EB-109: Model-Driven Biomedical Technology Development. Structural models of musculoskeletal tissues to calculate internal stresses and strains are an essential technology to investigate mechanisms of injuries and failure. Moreover accurate models provide understanding to predict the consequence of surgical interventions, device implantation, and tissue engineering. This proposal addresses the challenge of modeling the intervertebral disc. For over three decades finite element models have proved valuable to study disc mechanics. However, the currently available disc models have three significant limitations that will be addressed in this proposal for improved fidelity to predict mechanical behavior. First, the measured annulus fibrosus (AF) properties that are input into models are based on excised tissue tests. Excising the tissue disrupts the fiber structure, which introduces artifacts into the material property measurement and does not represent the inhomogeneities that are reflected by the gradual alterations in disc composition and microarchitecture. Second, excising tissue samples eliminates the interfaces between the sub-structures. Third, current disc models use "cable" elements that are embedded within an isotropic solid element for the AF. Ostensibly this represents the AF structure; however, the AF composition and microarchitecture is complex, requiring a constitutive formulation that better reflects the AF anisotropy, nonlinearity, and inhomogeneity. The objective of this proposal is to develop an image-based disc model that incorporates the tissue's inhomogeneous, anisotropic, and nonlinear material properties. We will obtain a 3D structural model with spatially distributed material properties using sophisticated enabling technologies. Aim 1: Extend the established 2D MRI to high resolution 3D MRI of the disc under load. Evaluate the performance of the pulse sequence and adjust the sequence parameters as required to optimize resolution, contrast, and signal-to-noise ratio. Aim 2: Image human discs under multi-axial loading configurations including compression, bending, and torsion and calculate 3D strain maps using state-of-the-art diffeomorphic image registration. Aim 3: Apply inverse methods to the established constitutive formulation and the measured strain maps to determine material properties. Validate the model and material properties by predicting strain under loading configuration not used in property calculations. Repeat the computational study using different loading modes for parameter regression. At completion of this study, we will be able to accurately predict the stresses and strains within the disc due to the microstructural contribution of the fibrillar components and extrafibrillar matrix and in response to interactions between the disc substructures. Determining material properties in situ is an innovative solution to the current limitations of artifacts in excised tissue tests and the disruption of sub-structural interfaces. The significance of this work lies in the development of an invaluable tool for accurate modeling of disc mechanics, which is central to both understanding mechanical failure of the disc and treating it. This system will initially be applied to cadaveric human tissues; however, long term advances in imaging may permit extension to in vivo application and patient-specific modeling and analysis. Economic impact The proposed experiments described in this Challenge Proposal will create four full time positions between the sites at the University of Pennsylvania (2.5 postdoctoral researchers, 0.2 technicians) and University of Minnesota (1 graduate student, 0.5 postdoctoral researchers). The University of Pennsylvania School of Medicine is an internationally recognized leader in the creation of new knowledge and therapies to improve human health, and in the training of the next generation of scientific leaders. In pursuit of these goals, the UPenn School of Medicine extends its economic impact widely and deeply throughout Pennsylvania, New Jersey, and Delaware. Recent studies attributed 37,000 jobs and $5.4 billion in economic activity to Penn Medicine in 2008. As Minnesota's only research university, the University of Minnesota is the economic engine for the state. As such the University of Minnesota's impact is felt far and wide, with more than 21,500 jobs directly related to research grants awarded to the University. This application addresses the broad Challenge area (06) Enabling Technologies, and the specific Challenge Topic, 06-EB-109: Model-Driven Biomedical Technology Development. Structural models that calculate the disc internal stresses and strains are essential to investigate mechanisms of disc injuries, tears, and failure. Moreover accurate models will provide understanding to predict the consequence of surgical interventions, device implantation, and tissue engineering. Unfortunately, the currently available disc models lack the fidelity to predict mechanical behavior accurately. The objective of this proposal is to develop an image-based disc model that incorporates the tissue's inhomogeneous, anisotropic, and nonlinear material properties. The proposed experiments described in this Challenge Proposal will create four full time positions between the sites at the University of Pennsylvania (2.5 postdoctoral researchers, 0.2 technicians) and University of Minnesota (1 graduate student, 0.5 postdoctoral researchers).