【作者】 郭超峰；

【导师】 张宏其；

【作者基本信息】 中南大学 ， 外科学， 2009， 博士

【摘要】 本研究应用计算机辅助工程（computer aided engineering,CAE）软件,建立了基于志愿者个体化CT图像的完整Lenke1BN型特发性脊柱侧凸三维非线性有限元模型,并对模型进行了有效性验证;在此基础上,仿真模拟Lenke1BN型特发性脊柱侧凸后路三维矫形手术,探讨选择不同下固定椎（lowest instrumented vertebra,LIV）的矫形效果。第一章Lenke1BN型特发性脊柱侧凸三维有限元模型的建立目的应用CAE软件,建立基于志愿者个体化CT图像的Lenke1BN型特发性脊柱侧凸三维非线性有限元模型。方法选择1例18岁女性Lenke1BN型特发性脊柱侧凸志愿者作为研究对象。取仰卧位,应用螺旋CT从T₁上缘至尾骨以1mm间距进行连续扫描,获得Dicom格式CT图像539张。导入逆向工程软件Mimics 10.01,建立包括胸-腰-骶尾椎和胸廓等结构的完整脊柱侧凸三维几何模型。对模型进行几何清理,然后导入有限元前处理软件HyperMesh 7.0划分有限元实体网格,并参照文献添加椎间盘及韧带结构单元,生成完整的特发性脊柱侧凸三维有限元模型。参照文献赋予模型材料参数和实常数,固定约束骶骨,在T₁上表面各节点沿Y轴施加100N拉力进行初步加载,检验模型的单元、节点质量能否达到计算要求。结果建立了完整的Lenke1BN型特发性脊柱侧凸三维非线性有限元模型,包括胸-腰-骶尾椎、椎间盘、胸廓及脊柱所有韧带、关节结构。模型共采用4种单元类型,14种材料性质;划分节点341228个,四面体单元1409929个,壳单元163132个,线缆单元715个和杆单元149个。对模型施加100N的拉力载倚,顺利通过计算。结论成功构建了基于志愿者个体化CT图像的Lenke1BN型特发性脊柱侧凸三维有限元模型,模型完整、逼真地还原了被模拟对象的脊柱特点;初步运算结果显示,有限元模型网格划分质量良好。第二章Lenke1BN特发性脊柱侧凸三维有限元模型有效性验证目的通过与志愿者临床实验以及历史体外（尸体）实验对比,验证所建立的Lenke1BN特发性脊柱侧凸三维有限元模型的有效性。方法（1）仰卧位左右Bending实验验证:在志愿者仰卧位左右Bending全脊柱X线上,测量T₁椎体中心偏移骶骨中垂线（centersacral vertical line,CSVL）的距离,分别设为A和B。对有限元模型进行如下约束、加载,模拟仰卧位左、右Bending实验:保持骶骨水平固定,约束和骶骨在X轴上平行的两边共4根肋骨最后方节点在Y轴方向的自由度,以此来模拟患者背部和摄片床的接触;在T₁中心节点,于X或-X方向施加载荷,使T₁中心相对CSVL分别向左、向右移动A和B大小的距离。比较X线片和有限元模拟凸侧Bending位下,各个弯Cobb角大小;并对比T₁-L₅各个椎体中心相对CSVL的偏移距离。（2）站立-仰卧实验验证:志愿者分别取站立和仰卧位拍全脊柱X片,并分别测量各个弯的Cobb角。根据志愿者体重,参照文献在有限元模型上分别模拟仰卧和站立,测量2种体位下各个弯的Cobb角,并和X片测量结果进行对照。（3）分段加载实验验证:在建立的模型中提取T_10-11、T₁₁-L₁以及L₁-S₁三个节段,分别参照历史同类体外（尸体）实验对有限元节段模型进行约束加载,并将加载结果与各自参照的体外实验结果进行比较,验证模型的有效性。结果（1）仰卧位左右Bending全脊柱X片上测量和有限元模拟测量,上胸弯、主胸弯与腰弯凸侧Bending位Cobb角分别为14°,26°,8°和15°,24°,7°;相比X线片测量,有限元模拟误差为8.3%。左侧Bending位X片测量,T₁-L₅各椎体中心至CSVL的平均距离为9.83±7.08cm,有限元模拟的平均距离为9.55±7.04cm,经配对t检验:t=1.77,p=0.095,按照检验水准α=0.05,判定二者没有统计学差异（P＞0.05）;右侧Bending位X片测量,T₁-L₅各椎体中心至CSVL的平均距离为10.15±7.34cm,有限元模拟的平均距离为10.02±7.35cm,经配对t检验:t=1.9002,p=0.0756,按照检验水准α=0.05,判定二者没有统计学差异（P＞0.05）。（2）站立位下,X片测量和有限元模拟测量上胸弯、主胸弯和腰弯的Cobb角分别为37°、50°、24°和33°、51°、24°;仰卧位下,X线测量和有限元模拟各个弯的Cobb角分别为29°、43°、22°和27°、42°、22°;相比X线片,有限元模拟平均误差为3.9%。（3）T_10-11、T₁₁-L₁和L₁-S₁各段有限元模型加载结果与各自参照的历史体外（尸体）实验结果基本吻合。结论从几何外形、仰卧位左右Bending实验、站立-仰卧实验及分段加载实验,验证了所建立的Lenke1BN型特发型脊柱侧凸模型的可靠性和有效性,为下一步生物力学模拟研究奠定了基础。第三章Lenke1BN型特发性脊柱侧凸后路三维矫形有限元模拟研究目的利用建立的Lenke1BN型IS有限元模型,模拟后路三维矫形手术,并探讨选择不同下固定椎对矫形效果的影响。方法应用建立的Lenke1BN型IS有限元模型,模拟后路全椎弓根螺钉固定三维矫形手术。具体约束加载如下:约束骶骨整体水平固定,参照文献在T₁-L₅各椎节分别施加模拟自身重力和肌肉因素的向下载荷,在固定节段凹侧模拟植入“椎弓根螺钉”,并放入“预弯”矫形钛棒,在棒末端施加向凹侧的旋转力矩,使棒向凹侧旋转90°,模拟旋棒矫形;同时在顶椎区(T₇-T₁₀)固定螺钉施加10Nm的扭矩,模拟椎体直接去旋转矫形。上固定椎选择T₄（上端椎+2）,下固定椎分别选择T₁₂（中立椎）、L₁（稳定椎）和L₂（稳定椎+1）,比较三种固定方案的矫形效果。结果顺利完成加载模拟矫形,选择T₁₂（中立椎）、L₁（稳定椎）和L₂（稳定椎+1）作为下固定椎模拟矫形后,上胸弯、主胸弯和腰分别矫正为:7.1°、7.4°、9.2°,6.4°、6.8°、8.3°和6.5°、7.2°、8.6°;矢状面胸椎后凸(T_5-12)分别为21.3°、20.7°和20.5°;三种矫形方案,矫形效果无显著差异。结论首次通过有限元模拟研究表明:对于中度Lenke1BN型特发性脊柱侧凸,选择性融合主胸弯可获得满意的腰弯自发矫正;应用全椎弓根螺钉固定结合顶椎区椎体去旋转技术,可将下固定椎从稳定椎上移至中立椎,减少远端融合节段。更多还原

【Abstract】 In current study,we established,based on CT images,and validated a complete three-dimensional finite element model of Lenke1BN idiopathic scoliosis（IS）,including all thoraco-lumbar-sacral vertebrae and thoracic cage,using computer aided engineering（CAE） softwares. On the basis of above,we simulated posterior three-dimensional correction surgery using this IS finite element model to investigate correction effectiveness with different lowest instrumented vertebra.Chapter One Establishment of Three-dimensional Finite Element Model of Lenke1BN Idiopathic ScoliosisObjective Using CAE softwares,to build three-dimensional finite element model of Lenke1BN idiopathic scoliosis based on CT images.Methods A 18-year-old female Lenke1BN idiopathic scoliosis patient was included as volunteer for current study.CT transverse scanning in supine position was done from T₁ to caudal end in 1mm layer interval,to obtain 539 CT dicom images.All CT images were imported into Mimics 10.01 to form qualified IS three-dimensional geometric model after geometry clean,including all thoraco-lumbar-sacral vertebrae and thoracic cage,which was further delivered to HypherMesh 7.0 to build 3d finite element IS model by mesh partition and quality control.A variety of material parameters were given to different mesh according to references,and 100N axial tension on top surface of T₁ was loaded for preliminary loading calculation using the finite element IS model,to check model quality.Results A three-dimensional finite element model of Lenke1BN idiopathic scoliosis was built successfully,including all thoracolumbar -sacral spine and thoracic cage,using 4 mesh types and 14 kinds of material parameters,in consist of 341228 nodes,1409929 tetrahedron elements,163132 shell elements,715 cable elements and 149 rod elements.Preliminary loading calculation was completed smoothly.Conclusions A three-dimensional finite element model of Lenke1BN IS in details,was built successfully based on individual CT images.Excellent preliminary loading calculation results using the FE model,proved mesh well-partitioned.Chapter Two Validation of Three-dimensional Finite Element Model of Lenke1BN Idiopathic ScoliosisObjective To validate 3-dimensional finite element model of Lenke1BN IS built in chapter one,by contrast with in vitro studies.Methods（1） Left and right Bending test validation:To measure the vertical distances between T₁ and center sacral vertical line（CSVL） on left and right supine bending X-ray film,finite element IS model was constrained and loaded to simulate supine bending test,with T₁ to be moved left and right the same distances to CSVL.The convex bending Cobb’s angles of proximal thoracic curve（PT）,main thoracic curve（MT） and lumbar curve（L）,and the average distance of T₁-L₅ to CSVL were compared between bending X-ray films and finite element simulation. （2） Erect-supine test validation:Finite element model were simulated erectly by applying corresponding downward forces to every segment for gravity and muscle actions according to references.Erect and supine Cobb’s angles of PT,MT and L were measured and compared on A-P X-ray films and finite element simulation.（3） Subsection validation: Segment T_10-11,T₁₁-L₁ and L₁-S₁ were extracted from the whole finite element model,and the three segments were respectively constrained and loaded referring to historical specimen biomechanical in vitro studies.Results（1） The convex bending Cobb’s angles of PT,MT and L curve on X-ray films and finite element simulation were 14°,26°,8°and 15°,24°,6°respectively.The error of finite element simulation was 8.3%. The average vertical distance of T₁-L₅ to CSVL on left and right supine bending X-ray film was 9.83±7.08cm and 10.15±7.34cm,with no significant difference comparing to finite element model simulation of 9.55±7.04cm and 10.02±7.35cm correspondingly（P>0.05）.（2） Erect Cobb’s angles of PT,MT and L were 37°、50°、24°and 33°、51°、24°on X-ray films and by finite element simulation.Supine Cobb’s angles of PT, MT and L were 29°、43°、22°and 27°、42°、22°on X-ray films and by finite element simulation.The average error of finite element simulation was 3.9%.（3） The segment simulation results were similar to their references respectively.Conclusions The three-dimensional finite element model of Lenke 1BN idiopathic scoliosis were well validated by geometry appearance, left and right supine bending test,erect-spuine test and segment validation,which was qualified for further biomechanical simulation study.Chapter Three Three-dimensional Finite Element Simulation of Posterior Surgical Correction of Lenke1BN idiopathic scoiliosisObjective To simulate posterior correction surgery using Lenke 1BN IS finite element model and investigate correction effectiveness with different lowest instrumented vertebra（LIV）.Methods Posterior pedicle screws on concave side and pre-bent rod placed were placed on IS finite element model,constraints and loadings were applied at the same time as follows:sacrum constrained horizontally,corresponding downward forces applied to every segment for gravity and muscle actions according to references,10Nm torsion moment against convex applied to apical zone,and proper torsion moment applied to rod to rotate 90 degrees backward.The upper instrumented vertebra（UIV）,was selected to T₄ and LIV down to T₁₂ （NV）,L₁（SV） and L₂（SV+1） respectively,with comparison of correction effectnesses among the three LIV choices.Results After simulations of posterior correction,Cobb’s angles of proximal curve,main thoracic curve,lumbar curve and T_5-12 in sagittal plane,fusion down to T₁₂,L₁ and L₂ were 7.1°,7.4°,9.2°,21.3°;6.4°6.8°,8.3°,20.7°and 6.5°,7.2°,8.6°,20.5°respectively.Correction rates were insignificantly different among the three LIV choices.Conclusion Selective thoracic curve fusion can lead to satisfactory spontaneous correction of lumbar compensatory curve for Lenke1BNIS. When all pedicle screws instrumentation and direct vertebra rotation of apical zone are used,fusion extended down to neutral vertebra（NV） is enough,with less segments fused than traditionally down to stable vertebra（SV）.更多还原