【作者】 米那瓦尔·阿不都热依木；

【导师】 阿达来提·阿合买提江；

【作者基本信息】 新疆医科大学 ， 劳动卫生与环境卫生， 2008， 博士

【摘要】 口腔颌面外科手术,尤其是正颌外科手术,主要针对面部骨性形态异常的患者,其目的不仅要改善功能,而且要恢复美观、悦人的外貌。为了获取最好的手术效果,需要在手术前选择一个最佳的手术方案。精细的手术要求外科医生与正畸医生的密切合作,以便制定出一个整合了临床检查、牙模及头影测量所获得的多方面信息的治疗计划。这需要医生能够在术前依据拟采取的术式、截骨部位、截骨量以及骨段移动的方向,预测患者术后的颜面外观及形态变化的结果,同时和患者及其亲友进行交流,再根据预测结果是否被认同最后确定和修正手术方案。多年来,术前无法预测术后面部软组织的形态变化一直是限制手术精确性的障碍。传统的方法只能通过头颅侧位片在二维空间上进行手术的设计和术后预测,预测结果不够准确。如何准确地预测术后的颜面软组织形态变化一直是临床医生关心和研究的课题之一。因此,对软组织形态变化进行逼真的预测是非常有必要的。目前已经有了应用计算机生成器官数字模型,并应用有限元分析手段来模拟手术的方法。计算机技术与医学的结合,尤其是与颅颌面外科的结合,促进了医学科学的飞速发展,在手术模拟、医学教育和培训、手术计划和术中辅助等方面都有许多应用,但寻求一种具有高度仿真性且便于临床应用的数字模型的方法仍有待于进一步探讨。本文采用三维螺旋CT扫描、三维重建、有限元手段及计算机模拟手术等方法,描述了颌面外科手术骨组织移位后面部软组织形态变化的预测方法,并建立了一种能较快速形成的患者头面部软组织有限元模型。由于在网格结构中纳入了部分面部肌肉,并根据其力学性能(各向异性、肌肉收缩的硬度)定义了肌肉及其周围结构,因而可较好地模拟肌肉运动下的面部软组织形变。这种方法形成的患者特异性头面部软组织有限元模型是适时、有效、并初步适合临床应用的。目的:创建正常人头面部软组织有限元模型,为正颌外科手术精细化提供理想的数字化工具及基础。在正常人头面部软组织有限元模型的基础上构建患者特异性头面部软组织有限元模型,并在患者特异性头面部软组织有限元模型上模拟手术,预测骨组织移位后软组织的形态变化情况。拟研究、开发和建立辅助正颌外科手术设计和预测手术效果的CAD系统。通过获取2例临床上拟行正颌外科手术的患者的医学信息并建立有限元模型,验证该系统的可行性和实用性。寻找解决颌面部手术术前难以预测术后面部软组织形态变化障碍的最佳途径。探索新的、更有效的颌面部手术的精确方法。开发正颌外科颅颌面三维立体可视化手术模拟系统,在计算机屏幕上完成正颌外科手术的实际操作过程,修订手术方案,进行医患交流并制定最佳手术方案。方法:1.应用有限元技术建立正常人头面部软组织模型的实验研究1.1正常人头面部三维螺旋CT扫描与重建选择1例发育正常、无错牙合畸形及骨性形态异常的正常志愿者1名,对该正常人的头面部进行三维螺旋CT扫描,扫描层厚/层距为0.625mm/0.625mm,获取头面部断层数据,扫描结果以DICOM格式保存。DICOM格式不能直接被有限元软件所应用,本文采用MIMICS10.01软件对正常人头面部骨组织与软组织两部分结构进行三维重建,重建间隔为0.25mm,采取自动提取轮廓与人工介入选择特征点的交互式轮廓提取方式分别构建骨组织、皮肤组织和皮下组织的三维几何模型,同时对构建的三维几何模型进行优化,最终以可直接被有限元软件所应用的STL文件格式保存。1.2正常人头面部软组织有限元模型的建立在CT扫描和重建结果的基础上,对已构建的骨组织、皮肤和皮下组织的三维几何模型采用手工划分的方式构建网格。设定颅面骨硬组织表面和边界条件,分割皮肤表面与颅面骨表面,对骨组织、皮肤和皮下组织均采用四面体(三角面片)的形式进行网格划分,采用3不同单元构成构建网格,并将6组面部肌肉纳入网格中;在网格结构中根据其力学性能对纳入的6组颌面部肌肉及其周围组织进行定义,同时应用有限元软件包ANSYS对划分的网格结构进行有限元分析,建立正常人头面部软组织有限元模型。由五位有经验的颌面外科和整形外科医生对比志愿者的外貌及CT资料,定性地分析构建的正常人头面部软组织有限元模型与志愿者外形的相似性,验证其临床实用性。探讨3种不同单元构成的网格在构建有限元模型方面的优势,为进一步构建患者特异性头面部软组织有限元模型奠定基础。2.患者特异性头面部软组织有限元模型的建立研究选择2例骨性下颌前突、拟进行正颌外科手术的患者,采用三维螺旋CT对患者的头面部软、硬组织结构进行CT扫描,并对输出的文件采用MIMICS10.01软件进行三维重建。按照最佳单元构成在患者的三维几何模型上分别进行骨组织、皮肤组织和皮下组织的网格划分。随后将构建的患者三维网格引入有限元软件包ANSYS中进行分析,导出2例患者的特异性头面部软组织有限元模型。同样由五位有经验的颌面外科和整形外科医生对比患者的外貌和CT资料,对2例患者特异性头面部软组织有限元模型进行定性的仿真分析。3.正颌外科手术的模拟及相应软组织形变的预测在构建的2例患者的特异性头面部软组织有限元模型中,按照资深正颌外科医生制订的手术方案进行手术的模拟及骨组织的移位。选定唇突点等数个软组织标志点作为软组织形变分析的基准。应用有限元软件包ANSYS以线弹性公式对这些软组织特征点位移的量及变化比率进行分析,确定骨组织移位后相应的软组织形变及其规律。由五位有经验的颌面外科和整形外科医生在2例患者的特异性有限元模型中定性地分析手术方案的可行性及精确性;根据颌面外科和整形外科医生的经验定性地分析对术后情况预测的可靠性,验证正颌外科手术中患者特异性头面部软组织有限元模型系统临床应用的可行性和有效性。结果:1.应用有限元技术建立正常人头面部软组织模型的实验研究1.1正常人头面部三维螺旋CT扫描与重建三维螺旋CT由于空间分辨率高,在头面部三维重建领域具有独特的优越性。它能一次快速地完成整个头面部医学信息的获取。本研究对正常志愿者的头面部进行扫描,共获取了465个断层。MIMICS 10.01(交互式医学图像控制系统)是可以显示和分割CT图像,对图像三维重建、渲染的交互式工具。它能完成头颅复杂曲面机构的重建,在医学领域里MIMICS软件可以用于诊断、手术计划或演习。本研究对获取的465个断层进行三维重建,分别获得了正常志愿者的骨组织、皮肤组织及皮下组织的三维几何模型,并对这3个三维几何模型进行优化,优化后的文件以STL格式保存,可直接被有限元软件应用,并进行有限元建模和一系列的有限元分析。1.2正常人头面部软组织有限元模型的建立对获取的医学信息采用手工划分网格的方式建立网格结构,其中的单元为四面体(三角面片)。通过三种不同单元构成的网格的对比,单元构成在10000至100000个单元之间的网格由于其误差较小、有限元分析计算时间适中,且便于肌肉的标记的特点,在构建有限元模型方面较其它两种单元构成的网格效果理想。本研究最终获得的正常人皮肤组织网格由36524个单元和18263个节点组成;皮下组织网格由38954个单元和19247个节点组成;骨组织网格由68722个单元和33689个节点组成。在获取的网格基础上,将对面部外形影响较大的6对肌肉纳入网格中,并以其功能不同进行了分组,并在有限元软件ANSYS中构建了正常人头面部软组织有限元模型。获取的有限元模型由65285个单元和32341个节点组成。该模型在一定程度上反映了面部软组织的生物力学性能(各向异性及肌肉的刚度),具有真实有效性,为建立患者特异性头面部软组织有限元模型创造了条件和基础。此过程需要大量手工网格划分及数学计算过程,耗时较长。应用有限元技术进行头面部软组织有限元模型的建立,对比正常志愿者的外貌和CT资料,定性分析的结果表明正常人头面部软组织有限元模型与CT扫描结果和外貌具有高度的相似性,可以在医院内/临床上推广,并可作为患者特异性头面部软组织有限元构建的基础。2.患者特异性头面部软组织有限元模型的建立研究本研究对2例骨性下颌前突患者的头面部进行扫描,共获取420个和375个断层。对获取的CT断层进行三维重建,分别获得了2例患者的骨组织、皮肤组织及皮下组织的三维几何模型,并对这3个三维几何模型进行优化,优化后的文件以STL格式保存在正常人头面部结构的网格划分和有限元建模方式的基础上,对获取的2例患者的医学信息采用手工划分网格的形式建立网格结构,其中的单元为四面体。病例1的皮肤组织网格由33686个单元和16812个节点组成;皮下组织网格由50848个单元和24990个节点组成;骨组织网格由个68690单元和33966个节点组成。病例2的皮肤组织网格由24196个单元和12092个节点组成;皮下组织网格由54038个单元和26807个节点组成;骨组织网格由61642个单元和30525个节点组成。在获取的网格基础上,将对面部外形影响较大的6对肌肉纳入网格中,并以其功能不同进行了分组,并在有限元软件ANSYS中构建了2例患者头面部软组织有限元模型。获取的病例1的有限元模型由76080个单元和37621个节点组成;获取的病例2的有限元模型由70410个单元和35009个节点组成。2个模型在一定程度上均反映了面部软组织的生物力学性能(各向异性及肌肉的刚度),具有真实有效性。两名患者的特异性头面部软组织有限元模型经五位有经验的颌面外科和整形外科医生验证与患者的外形有很高的相似性。3.正颌外科手术的模拟及相应软组织形变的预测在2例患者特异性头面部软组织模型上按照资深正颌外科医生制订的手术方案模拟正颌外科手术的截骨和骨移位,并预测骨组织移位后相应的软组织形变的情况。术后软组织的形态变化均以数据和图像两种形式输出。2例患者经手术模拟后的软组织特征点的位移最显著的是颏前点、颏下点、下唇突点及软组织B点,其次是上唇突点和软组织A点,其他标志点变化不大。应用所建立的正颌外科手术系统成功地完成了2例患者术后软组织形变的术前预测,手术模拟过程经五位有经验的颌面外科和整形外科医生验证具有很高的可靠性。软组织形变的预测与该五位有经验的外科医生的估计相符。此外,还可在该系统中对原手术方案进行修订,可取得满意的效果,具有很高的临床应用价值。结论:本实验应用有限元技术与计算机辅助颌面外科手术方法,在手工划分网格的基础上,在计算机上建立了一个能够反映部分软组织生物力学性能,包含皮肤、皮下组织并纳入肌肉组织的多层有限元模型;在正常人头面部软组织有限元模型建立的基础上构建了患者特异性头面部软组织有限元模型,同时在患者特异性头面部软组织有限元模型上进行了手术模拟和骨组织移位后相应软组织形变的预测。研究结果表明:1)本文不仅介绍了对正颌外科手术的一个完整方案的制定,同时着重于骨移位后面部软组织的形变,介绍了一种有限元模型框架,同时提供了一个可初步应用于临床的计算机辅助手术系统。2)本文提出了一个基于数个专家定性分析结果基础上的临床有效性方案。本课题为开发建立正颌外科手术模拟系统奠定了理论基础,该系统可作为临床正颌外科手术术前预测手术效果、修订手术方案的一种切实可行的方法,其临床应用效果自然逼真。3)这项技术不仅可用于下颌骨性前突患者,还可推广于上颌前突、偏颌畸形等其他骨性形态异常患者,同时还可以向颌面部复杂手术患者推广使用。在临床应用方面具有实际价值和推广意义。4)在2例临床病例中的应用均表明此模型系统可很好的适应患者外形,仿真性较高;软组织形变的预测与预期结果基本一致,可取得很好的临床效果。同时可应用于术前与患者的沟通和手术方案的修订,可在很大程度上提高正颌外科手术的精确性。可作为临床上制定和修订手术方案的有效手段。更多还原

【Abstract】 Maxillofacial surgery, especially orthognathic surgery, is addressed for patients suffering from maxillofacial dysmorphosis of the lower part of the face, i.e. from disequilibrium between the mandible, the upper jaw and the face. The purpose of the surgery is to improve the oral function, as well as the appearance of the patients. In order to sustain best operative result, a comprehensive surgical planning should be decided before the surgery. The delicate surgery requires close collaboration between surgeons and orthodontists, to develop a surgical planning that integrates multiple data gathered from different sources such as clinical examination (anthropometry), orthodontia (dental models) and radiology (cephalometry). It is important for the surgeons to predict the soft tissue deformation and the post-operative appearance of patients, according to the operative style, location, direction and amount of bone repositioning, to communicate with the patients and their family, and to revise the surgical planning. Not able to predict the soft tissue deformation resulting from bone repositioning before surgery is a major obstacle of the current orthognathic surgery. The traditional method, namely, teeth plaster casts, planning the surgical procedure and post-operative prediction on the basis of two dimensional space, is far from precise. How to precisely predict soft tissue deformation before the surgery is the subject that clinicians are concerning all these years along. Therefore, it is necessary to predict soft tissue deformation accurately. Computerized digital models using finite element method are currently created to simulate the orthognathic surgery by many researchers.The combination of computer science and medicine, esp. the cranio-maxillofacial surgery, promotes the development of medicine in many ways including simulation of surgery, medical education, training, surgical planning and computer-assisted surgery. Searching a clinically applicable method with highly accuracy needs further study.This paper addresses the prediction of facial soft tissue deformation resulting from bone repositioning in maxillofacial surgery, by means of SCT scanning, 3D reconstruction, finite element method, computer-assisted simulation. Finite element model of the facial soft tissue can be developed. Facial muscles are defined in the mesh as embedded structures, with different mechanical properties. Patient specific finite element models of 2 patients can be built on the basis of finite element modeling of a normal volunteer. Surgeries can be then simulated on the 2 patient specific models and the soft tissue deformation under muscle action can thus be performed. This patient model generation is robust, fast and easy to use, therefore seems compatible with clinical use.Objective:The purpose of this research is to create 3D finite element model of facial soft tissue. To create an ideal digital tool for delicate orthognathic surgical planning. To create an effective system for constructing the 3D finite element model of facial soft tissue. To create patient-specific model on the basis of 3D finite element model of normal volunteer. To simulate the surgical planning in the patient specific models. To create a system for predicting the soft tissue deformation and to apply the system in 2 cases to verify its accuracy and efficacy. To develop a CAD system for surgical planning and post-operative prediction of orthognathic surgery. To overcome the obstacles of orthognathic surgery and to investigate a more effective method for maxillofacial sugery. To develop a 3D visualization system for cranio-maxillofacial surgeries, by which surgeons could realize the surgical planning, revising, and the patient-doctor communication on the computer screen.Material and method:1.Finite Element modeling study of the facial soft tissue of a normal volunteer1.1 3D spiral CT scaning and reconstruction of the head and face of a normal volunteerContiguous spiral CT scaning with a scanning slice of 0.625mm/0.625mm to one normal volunteer, were generated on a GE Advantage CT scanning system. The CT data were saved in version of DICOM.The DICOM files could not be applied directly to finite element soft wares, MIMICS 10.01 soft ware was then used to reconstruct medical data. CAD models of skull, skin, and submucous tissue were reconstructed and optimized separately by MIMICS 10.01, and saved in version of STL.1.2 Development of 3D finite element model of the facial soft tissueThe boundary conditions were set and the mesh of the soft tissue and bone was manually made by using tetrahedrons on the basis of 3 CAD models of skull, skin, and submucous tissue. The biomechanical features were described in the mesh according to different anatomical structures. 6 pairs of facial muscles were included in the mesh. 3 different size of mesh were manually made to evaluate their advantage in developing the finite element model. Finite element model of the facial soft tissue was then developed by means of ANSYS soft ware package.Five surgeons qualitatively analyzed the conformity of the model to the morphology of the normal volunteer, by comparing the model with the appearance and the CT image of the volunteer.2.Patient specific finite element modeling study of the facial soft tissue of 2 patients with mandibular prognathismHead spiral CT scaning was also given to 2 patients who had disequilibrium between the mandible, the upper jaw and the face, which was more specifically mandibular prognathism. The 3 dimensional images were reconstructed with the CT data by using MIMICS10.01, and 3 CAD models of skull, skin, and submucous tissue were constructed. Mesh of skull, skin, and submucous tissue was manually made by using tetrahedrons. 2 patient-specific facial models were then developed by ANSYS soft ware package on the basis of the method of soft tissue modeling of the normal volunteer. Five surgeons qualitatively analyzed the conformity of the models to the patient morphology, by comparing the model with the appearance and the CT image of 2 patients.3.Surgical simulation and prediction of soft tissue deformation resulting from bone repositioningOrthognathic surgeries were simulated on 2 patient-specific face models according to the surgical planning made by experienced surgeons. Several anatomical soft tissue symbols were chosen to evaluate the soft tissue deformation with the linear elastic fomula by means of ANSYS soft ware package. The corresponding soft tissue deformation was simulated by analyzing the ratio and amount of all chosen soft tissue anatomical symbols.Five surgeons qualitatively analyzed the reliability of the simulated surgeries and the accuracy of the soft tisse deformation after surgery according to their own experience. The clinical applicability of 2 patient specific models was therefore verified.Result:1.Finite Element modeling study of the facial soft tissue of a normal volunteer1.1 3D spiral CT scaning and reconstruction of head and face of a normal volunteerBecause of the high contrast resolution, spiral CT has great advantages in acquiring medical data. This method can be widely used in hospital. Altogether 465 slices of head and face were acquired in our research.MIMICS10.01 (Materialise’s Interactive Medical Image Control System) is considered as an effective interactive digital tool revealing, segmenting and reconstructing CT images. It can quickly and clearly reconstruct the complicated structure of the whole head. MIMICS also offers approach for diagnosis, surgical planning in the field of Medicine. 3 CAD models of skull, skin, and submucous tissue were built and optimised by reconstructiong all these 465 slices from the 3D scanning. The acquired data were saved in the version of STL, which could be directly input into ANSYS package for modeling and analysis.1.2 Development of 3D finite element model of the facial soft tissueMesh of skull, skin and submucous tissue was constructed manually, which was combined with tetrahedrons. There were 36524 elements and 18263 nodes in the skin mesh, 38954 elements and 19247 nodes in the mesh of submucous tissue, and 68722 elements and 33689 nodes in the skull mesh. By constrast, the element size of 10000~100000 was considered best in our study, in developing finite element model for its less error, easier to label muscular structrure and proper caculating time.Finite element model of facial soft tissue was developed on the basis of mesh, 6 pairs of muscles which had major influence in the soft tissue deformation were included and grouped according to their fuction. The finite element model of facial soft tissue was combined with 65285 elements and 32341 nodes, which reflected the biomechanical features (anisotropy and stiffness depending on muscle contraction) in some respect and with certain accuracy. The model of facial soft tissues created the basis and condition for the development of patient specific soft tissue finite element models, which consumed great amount of manual work, computation and time.The facial finite element model was verified accurate qualitatively by 5 experience surgeons. The 3D finite element model of the facial soft tissue of the normal volunteer was proved highly conforming to the CT image, which could be widely used in the hospital in the future.2.Patient-specific finite element modeling study of the facial soft tissue of 2 patients with mandibular prognathism420 and 375 slices of head and face were acquired from 2 patients with disequilibrium between the mandible, the upper jaw and the face in our research. 3 CAD models of skull, skin, and submucous tissue were built and optimised by reconstructiong all these 420 and 375 slices of 2 patients from the 3D scanning by means of MIMICS 10.01 soft ware. The medical data were saved in the version of STL.Mesh of skull, skin and submucous tissue of 2 patients were constructed manually, which was combined with tetrahedrons. There were 33686 elements and 16812 nodes in the skin mesh, 50848 elements and 24990 nodes in the mesh of submucous tissue, and 68690 elements and 33966 nodes in the skull mesh of case 1. There were 24196 elements and 12092 nodes in the skin mesh, 54038 elements and 26807 nodes in the mesh of submucous tissue, and 61642 elements and 30525 nodes in the skull mesh of case 2.On the basis of the method building up the soft tissue finite element model of the normal volunteer, 2 patient specific face finite elment models were successfully built, which included 6 pairs of muscles and reflected the biomechanical features (anisotropy and stiffness depending on muscle contraction) in some respect and with certain accuracy. The patient specific finite elment model for case 1 was combined with 76080 elements and 37621 nodes, while the model for case 2 was combined with 70410 elements and 35009 nodes.2 patient specific finite element models were also qualitatively verified conforming to the CT image and patients’appearace. This patient specific model generation was proved to be robust, fast and easy to use. Therefore it seemed compatible with clinical use.3.Surgical simulation and prediction of soft tissue deformation resulting from bone repositioningBy applying the finite element system for orthognathic surgery, 2 orthognathic surgeries were successfully simulated, and revised before surgery on 2 patient specific finite element models according to the surgical planning made by experienced surgeons. The soft tissue deformations were also predicted. Results were outputed in form of both medical data and images. The most apparent soft tissue displacement was Mes. Menton of soft tissue, Pos. pogonion of soft tissue, and lower libial symbol, followed with soft tissue symbol A and upper libial symbol in both patient specific models. Its clinical applicability was also validated qualitatively by five surgeons according to their experience. The dimension, shape and anatomic contour of the facial models were considered similar to the patient morphology and precisely matched the surgical area. Both the patients and doctors were satisfied with the face models.Conclusion:This paper addressed to create a digital model of the facial soft tissue on the basis of finite element method and computer-assisted surgery, which was defined with biomechanical characteristics of facial soft tissue. The finite elment model reflected 2 layers of elements representing the skin, and submucous tissue seperatively. The patient specific models were developed on the basis of the 3D finite element model of the normal volunteer. Orthognathic surgeries were simulated on these patient specific models and the soft tissue deformation was predicted.The results of the research revealed:1.The paper not only introduced a full surgical planning of orthognathic surgery, but focused on the prediction of soft tissue deformation resulting from the bone repositioning also. A modeling frame and an effective method for developing patient specific finite element model were introduced.2.A clinical applicable protocol was addressed on the basis of qualitative analysis of 5 specialists. A CAD system for delicate orthognathic surgeries was developed, which could be a feasible method to predict surgical result and to revise surgical planning before the surgery.3.The system can not only be applied on mandibular prognathism, but also extended to maxillary prognathism, hemiprogathism and some other bone deformities, even to some complicated maxillofaical surgeries, which revealed great clinical value and significance.4.The patient-specific models fit the morphology of the patients, and the prediction of soft tissue deformation was considered coherent with what they would expect. Finite element method is proved to be helpful in computer-aided maxillofacial surgery, improving the accuracy of surgery in certain degree, which may be widely used in future.更多还原