【作者】 尤飞；

【导师】 胡庆夕；

【作者基本信息】 上海大学 ， 机械制造及其自动化， 2011， 博士

【摘要】 由于疾病、事故等原因所造成的大面积骨缺损,已经成为了临床医学治疗的重要难题之一。近年来,随着骨组织工程技术的迅速发展,利用种子细胞与仿生支架的复合体构建组织工程化骨,作为修复缺损骨的移植材料,已经成为骨缺损修复的一种全新的治疗模式,也成为了骨组织工程领域的研究热点。在组织工程化骨的构建中,仿生支架能够为新生的骨组织提供合适的生长空间和足够的机械支撑,并能够介导细胞间的信号传递和相互作用,诱导新骨的形成。因此,仿生支架的制备成为骨组织工程研究的关键,具有极为重要的研究价值。仿生支架的制备包括解剖外形的制造和内部微观孔的构建。传统的制备方法更多地致力于内部微观孔的构建,解剖外形主要依靠手工成型或模具成型,制造精度相当低。而且,支架内部结构性能无法在制造前进行评价,孔隙率、连通性等参数在制造过程中也难以控制。随着快速成形技术在骨组织工程领域应用的不断深入,使得利用快速成形方法制备仿生支架成为一种更为理想的途径。采用快速成形技术,不仅可以一次成型出精确的支架解剖外形和复杂的内部微观孔结构,而且可以根据支架的三维模型进行内部结构性能的加工前评价,从而制造出性能优良、结构合理的仿生支架。因此,仿生支架建模是利用快速成形方法制备仿生支架的一项关键技术。正是基于上述的研究背景和应用需求,本文以仿生支架为研究对象,面向骨组织工程领域展开了仿生支架的外观模型和内部结构模型的构建,以及支架内部结构性能评价等方面的研究工作。主要的研究内容及成果包括以下几个方面:(1)基于Delaunay三角剖分算法思想,提出了缺损骨曲面模型的重建方法。根据CT图像数据构建像素立方体,通过骨组织图像的灰度阈值的比较,实现空间点云数据的采集。运用K-近邻求解和二维Delannay近邻算法,简化了空间点集的Delannay三角化过程,加速了曲面重建的效率。利用局部可变球映射算法解决了孔洞特征拓扑重构的问题,从而获得精确的缺损骨曲面模型。同时,重构的曲面模型为仿生支架的解剖外形以及内部微观孔结构的构建,提供了良好的数字模型。(2)在分析骨缺损的病理特征和种类的基础上,提出了面向骨缺损修复的仿生支架外观模型的构建方法。该方法采用二面角判别算法自动搜索孔洞的边界,获取孔洞边界边所构成的三维多边形;基于面积最小原理进行多边形的三角剖分,根据边长最短原则进行网格细化,实现三维多边形的Delaunay三角剖分;利用加权伞算子控制修补曲面的曲率变化,进行曲面网格的光顺处理,保证修补曲面片与周围曲面的光滑过渡。通过各模型之间的布尔运算构建仿生支架的外观模型,大大地降低建模的复杂度。(3)采用多约束背包问题模型与混合遗传算法综合的方法来构建仿生支架的内部微观孔结构模型。基于多约束背包问题模型,采用椭球体作为单元体,构建支架内部微观孔结构的负模型,利用混合遗传算法进行求解。在遗传算子设计中,采用比例选择与最优保存策略相结合的混合选择算子,提高算法的运行效率和收敛性;采用均匀交叉算子避免种群多样性的退化,使得微观孔结构呈多样性;采用均匀变异算子增强算法的局部搜索能力,促进群体的多样性演化;采用扰动算子对解空间进行局部调整,提高算法的全局搜索能力。然后,通过支架的外观模型与负模型之间的布尔运算,获得含微观孔的仿生支架模型。(4)分析了影响仿生支架内部结构性能的因素,建立了支架微观孔结构性能的评价指标体系,包括孔隙率、连通性、均匀性、扭曲度和比表面积五项评价指标。基于支架微观孔结构的负模型,提出了各项评价指标值的计算方法。分析了各项评价指标对支架的生物活性、力学强度、降解速度等性能的影响程度,基于AHP方法确定了各项评价指标在支架结构性能综合评价中的权重值。基于灰色关联度分析的评价理论,提出了仿生支架内部结构性能的综合评价模型,并通过计算各项评价指标的灰色关联度来综合评判支架内部结构性能的优劣。(5)根据上述的理论和方法,采用面向对象技术和可视化技术开发了仿生支架建模的原型系统,初步实现了缺损骨曲面模型的三维重建、修复体模型的构建、支架内部结构建模及性能评价等功能,验证了所提出的建模方法的科学性、合理性和正确性。更多还原

【Abstract】 The large defect of bone is one of the key problems of clinical medicine treatment, due to disease, accident, and so on. The technology of bone tissue engineering is developing quickly in recent years. A new treatment pattern of bone defect repairing is that the defective bone is repaired with tissue-engineered bone as transplant, and the tissue-engineered bone is generated by co-combining with cultured osteogenic cells and bionic scaffold in vitro. This is a hotspot in the research field of bone tissue engineering too. The bionic scaffold provides suitable growth space and enough mechanical support for regenerating bony tissue, mediates the signal and the interaction between osteocytes, and induces the regeneration of osteocytes in the tissue-engineered bone regeneration. Therefore, the fabrication of bionic scaffold is the key technology of bone tissue engineering and most important research.The fabrication of bionic scaffold includes the fabrication of anatomical shape and the generation of inner porous structure. The traditional method focuses much more on the generation of inner porous structure, and anatomical shape is fabricated by die or handcraft. The precision of scaffold is fairly low. Moreover, the performance of inner structure of scaffold can not be evaluated before being fabricated, and some parameters, such as porosity and connectivity, are difficult to be controlled. With the continuous development of applying the technology of Rapid Prototyping (RP) to the field of bone tissue engineering, a better method to fabricate bionic scaffold using RP is set up. Not only precise anatomical shape and complicated inner porous structure of scaffold can be synchronously generated by using RP, but also the performance of inner structure of scaffold can be evaluated on the basis of three-dimensional model information before being fabricated. Therefore, the modeling is a key technology of fabricating bionic scaffold by using RP.According to the research background and application requirement above, bionic scaffold is chosen as research object in this paper, and a series of research oriented to bone tissue engineering are developed, including modeling of anatomical shape and inner structure of scaffold, performance evaluation of inner structure, and so on. The main research contents and contributions are described as following.Firstly, based on the theory of Delaunay triangulation, three-dimensional reconstruction method of surface model of defective bone is set up. The pixel cube model is constructed based on data of CT images. The data of space point cloud is acquired by comparing the gray threshold of image of bone tissue. The algorithm of k-nearest neighbors and 2D-Delaunay neighbors are applied to simplify Delaunay triangulation of space point set and accelerate the surface reconstruction. It is better to adopt the algorithm of local deformable spherical map to solve the problem of topological structure reconstruction of holes, and then the precise surface of defective bone is generated. Moreover, the reconstructed surface provides a more precise digital model for modeling of anatomical shape and inner porous structure of bionic scaffold. Secondly, on the basis of analyzing clinicopathological characteristics and type of bone defect, the modeling method of shape model of bionic scaffold is set up. The boundary edge of hole is automatically searched to generate 3D polygons by adopting the dihedral angle criteria algorithm. Delaunay triangulation of 3D polygon is finished based on the minimum area principle, and the mesh is refined based on the shortest edge principle. The weighted umbrella-operator is applied to control the curvature transformation of the patching mesh to smooth it, and the fairing surface patch is merged into the surface around it. Then, the shape model of bionic scaffold is generated by Boolean operation between every two key models. Thus, the complexity of modeling is greatly reduced.Thirdly, the modeling method of inner porous structure of bionic scaffold is set up by combining the structure of multi-constrained knapsack problem model with hybrid genetic algorithm. Based on the structure of multi-constrained knapsack problem model, and using ellipsoid as the basic unit, the inverse model of porous structure is generated, and hybrid genetic algorithm is used to solve this model. In the design of genetic operators, the hybrid selection operator of combining proportional selection with optimal Elitist Model is adopted to improve the operating efficiency and the convergence. The uniform crossover operator is adopted to avoid the diversity deterioration of population, and then to improve the bio-diversity of porous structure. The uniform mutation operator is adopted to enhance the local searching ability of hybrid genetic algorithm, and then to promote the bio-diversity evolution of population. The perturbation operator is adopted to locally adjust the solution space, and then to enhance the overall searching ability of hybrid genetic algorithm. Then, the model of bionic scaffold containing porous structure is generated by Boolean operation between the shape model and the inverse model.Fourthly, according to factors influencing the performance of inner structure of bionic scaffold, the evaluation index system of porous structure performance of scaffold is constructed, including porosity, connectivity, uniformity, twist degree and specific surface area. The calculation method of every evaluation index is set up based on the inverse model of porous structure of scaffold. Every evaluation index influencing bioactivation, biomechanical strength and degradation rate is analyzed, and the weight value of every evaluation index is calculated based on the theory of AHP. The integrated evaluation model of inner structure performance of bionic scaffold is constructed based on the theory of gray relation grade, and the gray relation value of every evaluation index is calculated to synthetically evaluate inner structure performance of bionic scaffold.Lastly, according to the theory and method above, the modeling prototype system of bionic scaffold is designed by adopting object-oriented technology and visualization technology. This system can perform three functions, including three-dimensional reconstruction of surface model of defective bone, prosthesis modeling, inner structure modeling of bionic scaffold and evaluation of its performance. The scientificity, rationality and validity of the described modeling theory and method are validated.更多还原