不同材料属性分配方法对椎体有限元模型力学性能的影响

发布时间：2018-06-25 04:20

本文选题：有限元分析 + 脊柱　；参考：《南方医科大学》2013年硕士论文

【摘要】：背景：数字医学概念的提出促进了有限元技术与医学的结合,CT数据集在三维重建方面的广泛应用更加快了这一进程的发展,尤其是在骨骼模型的构建方面显得特别突出。三维有限元模型可以用于正常骨骼的应力应变分析,也可用于分析其发生骨折的危险系数或内固定物的微动情况。在国外,O'Reilly和Whyn, Schmidt等,Little等,Schileo等,Yosibash等对脊柱和股骨等部位进行了相关的研究。在国内有限元方面研究也逐渐加深,特别是数字医学概念的提出更是把有限元技术与医学方面结合起来。很多学者在有限元仿真等领域做出相应的研究。有限元建模是指建立一个可以由计算机顺利计算的数字模型,该模型要和现实中物理模型的力学性能相一致,其实质是用有限元方法在计算机的虚拟环境中建立一个数字模型,这个模型有以下要求：一,要保证力学的完整性；二：要保证计算的有效性。即建立的有限元模型有两个使命,对上要承载完整的力学信息,对下要保证计算机可以快速准确的计算。在建立有限元模型的过程中既要保证所建模型外形与现实相近,也要使模型的一些力学材料性能与真实情况尽量相符。Marom等认为CT扫描可以反映出骨骼模型的高特征性。CT扫描作为一种临床常规体外非侵入性的诊断方法,它不仅可以提供骨骼精确的几何学信息,更可以提供骨骼相关的机械性能参数。现阶段,CT数据在有限元领域的最主要用途在于骨骼三维模型构建,而对于有限元模型构建的另一个重要组成部分：材料属性赋值,却甚少涉及。骨骼是医学有限元分析的主要对象之一,人骨的材料属性更是十分复杂的,个体差异明显,不同人或同一人体不同部位之间都存在差别。传统的骨骼材料属性划分方法是将骨单纯分为皮质骨和松质骨两部分进行赋值,这样划分必然与人骨材料属性复杂性不符,在计算中不免会出现较大误差。随着人们对CT数据认识的加深和相应软件的开发运用,逐渐开发出的一类软件如MIMICS, BONEMAT, AMAB等,这些软件可以把CT数据作为一种内部的数据代码与有限元仿真模型相结合,其分配方案是根据CT数据的灰度值来决定有限元模型的材料属性的,具体方法是通过提取出单元体积内所有CT图像的像素值,并计算出其平均值来赋予到该单元中。这种方法可以最终达到通过CT灰度值确定有限元模型中每个单元弹性模量的目的。医学图像的三维重建及可视化技术是一种运用计算机图形学、图像处理、计算机视觉以及人机交互技术将医学图像数据转换为图形或图像在屏幕上显示出来,并进行交互处理的理论、方法和技术。MIMICS是Materialise公司研发的交互式的医学影像控制系统(Materialise's interactive medical image control system),是一套高度整合而且易用的3D图像生成及编辑处理软件,它能输入各种医学影像(CT、MRI)数据,建立3D模型进行编辑,在PC机上进行大规模数据的转换处理。该软件除了可以运用于构建3D模型,更可以将CT灰度值与模型的材料属性联系起来,按照一定的划分梯度对模型的材料属性进行赋值。一个较好的赋值方法,配合一个较好的分配方案,可以改善计算中因个体差异造成的误差。本实验就是在MIMICS软件的基础上在完成较好的几何重建的前提下,运用其材料属性分配模块对模型的弹性模量按照不同分配梯度进行赋值,并将这些赋值后模型的有限元分析结果进行比较,再将其与手动赋值方法模型有限元结果比较,得出较好的赋值方法。通过研究两种赋值方法(CT灰度值赋值法、手动赋值法)及CT灰度值赋值法中不同梯度变化对有限元分析结果的影响,得出一个相对经济、准确的划分梯度,为将该法在有限元研究中的应用特别是临床相关快速建模分析等方面提供相关的理论依据。椎骨有限元建模本部分内容重点对课题实施的前期建模工作进行归纳总结,阐述了相关的方法和技术,细致的说明了本实验进行有效建模的全过程。有限元分析是近年来在生物力学领域仿真人体结构力学功能的一个重要实验手段。通过建立人体有限元分析模型,赋予模型相应的材料属性并合理模拟在体条件,可以有效地分析人体结构的物理性质。其中,建立有限元几何模型是有限元仿真分析的基础,其目的就在于为进一步的有限元分析过程提供一个物理框架,然后可以对这个框架进行相应的修饰完善,最终达到可以完成分析的目标。本部分内容就是以胸腰椎标本为例来说明相应骨骼有限元模型建立的一般流程技术方法及注意事项。重点阐述了对三维重建中构造几何模型的过程,通过相应的Remesh操作提高有限元网格划分成功率及质量的方法,以及使用geomagic软件处理相应三维模型得到理想有限元模型的方法。自动分配方法中材料属性分配梯度对椎骨有限元模型力学性能的影响 [目的] 研究以灰度值为基础的不同材料属性分配梯度对椎骨有限元模型力学性能的影响。 [方法] 对一位健康成人脊柱(T12～L5节段)进行快速CT薄层扫描,在MIMICS中对每节椎骨进行三维重建,通过Geomagic软件处理后,导入ANSYS中进行网格划分,再返回到MIMICS中按2,4,8,10,50,100,200,400份等8种梯度对材料属性进行分配,最后重新导入ANSYS软件按照同一载荷条件进行有限元分析。 [结果] 2、4、400这三个分配梯度与其它分配梯度相比,应力情况存在显著性差异(P0.05),而8、10、50、100、200梯度之间计算结果偏差不大。 [结论] 有限元模型材料属性划分不宜过多或过少,10份左右的划分梯度既可保证运算结果的精确性,也相对可以提高运算速度,尤其适用于临床个性化快速有限元建模。材料属性手动分配法与自动分配法对椎骨有限元模型力学性能的影响 [目的] 研究自动分配方法与手动赋值法对椎骨有限元模型力学性能的影响。 [方法] 对一位健康成人脊柱(T12-L5)进行快速CT薄层扫描,在MIMICS中对每节椎骨进行三维重建,通过Geomagic软件处理后导回MIMICS中进行Remesh处理,然后导入ANSYS中进行网格划分。材料属性按两种方法划分：一、手动赋值法,直接在ANSYS中选取外围单元赋予皮质骨属性,其它单元赋予松质骨属性；二、CT灰度值赋值法(自动分配法),模型导回到MIMICS中按照10份得分配梯度进行材料属性的赋值。最后在ANSYS中进行有限元分析。 [结果] 手动赋值法中椎体的最大位移为28.0583mm,最大应力为149.167N；CT灰度10分法椎体的最大位移处位移为24.3517mmm,最大应力为148.986N。节点的位移分布情况,手动赋值法中位移活动范围等值线较少；CT灰度10分法呈多处等值线。根据路径图可以看出.两种方法运算结果存在显著性差异。直接赋值法较灰度赋值法应力情况总体偏低,其中皮质骨部分应力显著偏大,松质骨部分应力明显偏小。 [结论] 手动赋值法的应力分布数据在皮质骨和松质骨之间分布相对离散,CT灰度值赋值法的应力分布数据分布相对集中。相比较而言灰度赋值法10分法其分法更贴近于人骨复杂材料属性的状况,尤其适用于临床个性化快速有限元建模。
[Abstract]:Background:
The concept of digital medicine has promoted the combination of finite element technology and medicine. The extensive application of CT data set in 3D reconstruction has accelerated the development of this process, especially in the construction of bone model. The three-dimensional finite element model can be used to analyze the stress and strain of normal bone, and can also be used to analyze it. The risk factor of fracture or the micromovement of internal fixation. In foreign countries, O'Reilly and Whyn, Little et al, Schileo et al, Yosibash and other parts of the spine and femur have been studied. The research on the finite element in China is gradually deepened, especially the concept of digital medicine is put forward by the finite element technology and the medical prescription. Many scholars have made corresponding research in the field of finite element simulation.
Finite element modeling refers to the establishment of a digital model that can be successfully calculated by a computer. The model is consistent with the mechanical properties of the physical model in reality. The essence of this model is to establish a digital model in the virtual environment of the computer by the finite element method. The model has the following requirements: first, to ensure the integrity of the mechanics; two: to The validity of the calculation is guaranteed. That is, the established finite element model has two missions to carry the complete mechanical information to ensure that the computer can calculate quickly and accurately. In the process of establishing the finite element model, we should not only ensure that the shape of the model is similar to the reality, but also the performance of some mechanical materials and the real situation of the model. The high characteristic.CT scanning that CT scan can reflect the bone model can be used as a non invasive diagnostic method of clinical routine in vitro. It can not only provide accurate geometry information of bone, but also provide bone related mechanical properties. At this stage, the most important use of CT data in the finite element field is at the present stage. The construction of three-dimensional model of bone, and another important part of the construction of finite element model: material attribute assignment, but little involved. Bone is one of the main objects of medical finite element analysis, the material properties of human bone are more complex, individual differences are obvious, there are differences between different people or different parts of the same human body. The traditional method of dividing the properties of bone materials is to assign the bone to two parts of the cortical bone and the cancellous bone, which is incompatible with the complexity of the human bone material property, and there will be great errors in the calculation. With the deepening of the understanding of the CT data and the development and application of the corresponding software, a class of software developed gradually. Such as MIMICS, BONEMAT, AMAB and so on, these software can combine the CT data as an internal data code and the finite element simulation model. The allocation scheme is based on the gray value of the CT data to determine the material properties of the finite element model. The specific method is to extract the pixel values of all the CT images in the single volume and calculate it. The average value is assigned to the cell. This method can finally achieve the purpose of determining the elastic modulus of each element in the finite element model through the CT gray value.
3D reconstruction and visualization of medical images is a theory that uses computer graphics, image processing, computer vision and human-computer interaction technology to convert medical image data into graphics or images on screen and interact with each other. Method and technology.MIMICS is an interactive medicine developed by Materialise company. The Materialise's interactive medical image control system is a highly integrated and easy to use 3D image generation and editing processing software. It can input various medical images (CT, MRI) data, establish 3D model for editing, and transform the large-scale data into the PC machine. This software can be used in addition to the application of the software. In building the 3D model, we can connect the CT gray value with the material properties of the model, and assign the material properties of the model according to a certain partition gradient. A better assignment method and a better allocation scheme can improve the error caused by individual difference in the calculation. This experiment is based on the MIMICS software. On the premise of better geometric reconstruction, the modulus of elasticity of the model is assigned by its material attribute allocation module, and the results of the finite element analysis of these models are compared, and a better assignment method is obtained by comparing it with the finite element method of the manual assignment method. Two kinds of assignment methods (CT gray value assignment method, manual assignment method) and the influence of different gradient changes on the finite element analysis results in the CT gray value assignment method, a relative economic and accurate division gradient is obtained, which provides the relevant theoretical basis for the application of the method in the finite element study, especially in the rapid modeling and analysis of clinical phase.
Finite element modeling of vertebrae
This part mainly summarizes the early modeling work of the implementation of the subject, expounds the related methods and techniques, and illustrates the whole process of effective modeling in this experiment. Finite element analysis is an important experimental means to simulate the function of human structure and mechanics in the field of biomechanics in recent years. The finite element model is the basis of the finite element simulation analysis. The aim of this model is to provide a physical framework for further finite element analysis, and then the frame can be applied to this frame. In this part, the general flow technique method and attention for the establishment of the corresponding skeleton finite element model are illustrated by the example of the thoracic and lumbar specimens. The process of constructing the geometric model in the three-dimensional reconstruction is emphasized, and the corresponding Remesh operation is used to improve the process. The method of limiting the success rate and quality of mesh generation and the method of obtaining the ideal finite element model by using Geomagic software to process the corresponding three-dimensional model.
Effect of gradient of material property distribution on mechanical properties of finite element model of vertebrae in automatic allocation method
[Objective]
The effects of different material property distribution gradients based on gray value on mechanical properties of vertebral finite element models were studied.
[method]
A healthy adult spinal column (T12 ~ L5 segment) was scanned by fast CT thin layer, and each vertebra was reconstructed in MIMICS. After Geomagic software, the mesh was introduced into ANSYS and then returned to MIMICS to distribute the material properties according to the 8 gradients such as 2,4,8,10,50100200400, and finally reimported the ANSYS software. The finite element analysis is carried out according to the same load condition.
[results]
Compared with other distribution gradients, there is a significant difference in stress between the three distribution gradients of 2,4400 (P0.05), but the results of the 8,10,50100200 gradient have little deviation.
[Conclusion]
The material attributes of the finite element model should not be too much or too little. The 10 partition gradient can not only ensure the accuracy of the calculation results, but also improve the computing speed. It is especially suitable for the personalized rapid finite element modeling of clinical personalization.
Effects of manual distribution and automatic allocation of material properties on mechanical properties of finite element models of vertebrae
[Objective]
The influence of automatic allocation method and manual assignment method on the mechanical properties of the finite element model of vertebrae is studied.
[method]
A healthy adult spinal column (T12-L5) was scanned by fast CT thin layer, and each vertebra was reconstructed in MIMICS. Remesh treatment was carried out in MIMICS through Geomagic software, and then transferred into ANSYS to mesh. The material attributes were divided according to two methods: first, manual assignment method, directly selecting peripheral single in ANSYS. The element gives the cortical bone attribute, the other units give the cancellous bone property; two, the CT gray value assignment method (automatic distribution method), the model is guided back to the MIMICS to assign the material attribute according to the distribution gradient of 10 parts. Finally, the finite element analysis is carried out in the ANSYS.
[results]
The maximum displacement of the vertebral body in the manual assignment method is 28.0583mm, the maximum stress is 149.167N, the maximum displacement of the CT gray 10 point method is 24.3517mmm, the maximum stress is the displacement distribution of the 148.986N. node, the displacement range contour line is less in the manual assignment method, and the CT grey degree method has multiple contour lines. We can see that there are significant differences in the results of the two methods. The stress in the direct assignment method is generally lower than that of the gray value method. The partial stress of the cortical bone is significantly larger and the partial stress of the cancellous bone is obviously smaller.
[Conclusion]
The distribution of stress distribution in the manual assignment method is relatively discrete between the cortical bone and the cancellous bone, and the distribution of the stress distribution data of the CT gray value assignment method is relatively concentrated. In comparison with the gray value method, the 10 division method is more close to the state of the human bone complex material properties, especially for the personalized rapid finite element modeling of clinical individualization.
【学位授予单位】：南方医科大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：R318.08

【相似文献】