Multimodal image analysis and modeling of thin bone structures in the human skeleton

人体解剖学的计算机建模是理解结构和功能的有力途径。有限元(FE)方法能够表示肌肉骨骼系统的几何和材料复杂性，特别是在由3D医学成像数据生成的模型中。实验验证表明，有限元分析可以模拟简化载荷和边界条件下的肌肉骨骼力学行为。尽管临床CT成像存在分辨率限制，但我们开发的图像处理方法可以准确反映特定样本FE建模所需的薄骨几何形状和密度。这与高度自动化的工作流程相结合，使得可以生成多个有限元模型来量化这些复杂结构的力学行为。然而，要表示人类肌肉骨骼解剖的生理行为，需要复杂的载荷和边界条件。此外，即使在发生故障后，了解这些结构的力学行为也是至关重要的。本研究试图将多模式图像分析和有限元建模相结合，以准确地描述在复杂生理载荷和骨折后稳定性下薄骨结构的力学行为。在这项工作中，我们建议解决以下具体目标： 1.集成通过医学成像和肌肉骨骼建模模拟识别的多模式信息，以更好地实现薄骨骼结构中的生理负荷和边界条件的域，特别是颅颌面骨骼(CMF)和骨盆。 2.利用实验设计方法量化薄骨模型对载荷和边界条件的敏感性，以识别关键的骨折风险情景。 3.利用序列图像分析评价薄骨结构屈服后的稳定性，建立能够反映有限元分析中破坏和压实的本构模型。 最终，这项研究旨在建立一个健壮的、经过实验验证的平台，提供对人类薄骨结构的生理力学行为的准确表示。这对于了解损伤和疾病对薄骨结构的影响，利用新的和现有的技术进行修复或再生具有重要的应用价值，并将推动在生物医学研究和设计中继续使用有限元建模。

Computer modeling of the human anatomy is a powerful approach to understanding structure and function. The finite element (FE) method enables representation of geometric and material complexities of the musculoskeletal system, specifically in models generated from 3D medical imaging data. Experimental validation has demonstrated that FE analysis can model musculoskeletal mechanical behaviour under simplified loading and boundary conditions. We have developed image processing methods that accurately reflect thin bone geometry and density necessary for specimen-specific FE modeling, despite resolution limitations imposed with clinical CT imaging. This, combined with highly automated workflows, has allowed the generation of multiple FE models that quantify the mechanical behaviour of these intricate structures. Yet, to represent physiologic behaviour of the human musculoskeletal anatomy requires complex load and boundary conditions. Further, understanding the mechanical behaviour of these structures is critical even after failure has occurred. This research seeks to integrate multimodal image analysis and FE modeling to accurately depict the mechanical behaviour of thin bone structures under complex physiologic loading and post fracture stability. In this work we propose to address the following specific aims: 1. To integrate multimodal information identified through medical imaging and musculoskeletal modeling simulations to yield an improved realization of the domain of physiologic load and boundary conditions in thin bone structures, specifically the craniomaxillofacial skeleton (CMFS) and pelvis. 2. To quantify the sensitivity of thin bone models to load and boundary conditions utilizing a Design of Experiments approach to recognize critical fracture risk scenarios. 3. To evaluate post yield stability in thin bone structures utilizing sequential image analysis and create constitutive models which can represent failure and compaction in FE analysis. Ultimately this research aims to establish a robust experimentally validated platform that provides accurate representation of the physiological mechanical behaviour of human thin bone structures. This has important application to understanding the impact of injury and disease on thin bone structures, the utilization of new and existing technologies for their repair or regeneration and will motivate the continued use of FE modeling in biomedical research and design.