【作者】 刘清华；

【导师】 余斌；

【作者基本信息】 南方医科大学 ， 骨外科学， 2010， 博士

【摘要】 研究背景：随着现代化的进程,交通业和建筑业空前发展,以及人们健康观念的提高,足踝部疾病发病率逐年上升,病种也日趋复杂,使骨科临床面临新的挑战,迫切需求对其进行深入研究。足是人体承重的根本,直接与地面接触,除了行走功能外,还能够吸收运动振荡、缓冲地面反作用力,保护躯体重要器官和组织免受伤害,并使人体的重心稳定。踝关节是人体与地面接触的枢纽,行走、跳跃、跑步和登高都需要踝关节的参与,即便骑自行车或驾驶汽车亦离不开踝关节的协调动作,可以说日常生活中的每一个动作都有踝关节的参与,因此踝关节也是最容易受到损伤的关节之一。踝关节力学机制复杂,各种损伤后都可能打破其周围结构的力学平衡而导致不稳定,诱发创伤性关节炎。而且,足踝部结构异常除导致疼痛、畸形及活动障碍外,还可进一步影响下肢、骨盆、脊柱等力学功能的正常发挥,故对其研究具有重要临床意义。足踝部生物力学的研究是人体生物力学研究中的重要组成部分。许多研究者指出生物力学因素在探求足踝部疾病的病因机制、治疗和预防等方面起到重要的作用。以往对足踝部研究多采用传统实验生物力学测试的手段,但是用这种方法的缺点在于实验手段复杂,进行多种载荷工况下的实验往往耗资大、周期长、效率低,而且由于无法在人体上直接实验,因此很难准确地反映出各种不同受力情况下的应力分布规律。为了解决这一问题,近年来,随着数字化技术的提高,越来越多的研究者采用数值计算的方法进行分析,即以传统的力学分析理论为基础,应用数值分析手段例如有限元方法(finite element method, FEM)等,进行线性和非线性的应力和变形分析。利用FEM可以仿真模拟各种足踝部疾患,从而使得对足踝部复杂的骨胳几何结构、边界条件和材料的不均匀性等问题的生物力学研究有了可能的解决途径。FEM具有强大的建模功能,在动静状态下能够对具有复杂的几何形状、材料参数和不同受力条件下的物体进行模拟仿真研究,其已经越来越多的被应用到人体生物力学中。因足踝部骨、关节、韧带、肌腱等结构复杂,在进行有限元建模时工作量较大,且缺乏完善的各种组织详细参数,有限元模拟相对于其他大关节的研究难度更大。现有建立的足踝部有限元模型,大部分对踝关节的解剖结构建立不够完善,主要用于与足底压力布相关的研究。然而踝关节损伤对人体功能的影响更加受到临床的重视,亟待对其更深入地研究。本研究是在总结以前研究的基础上,对足踝部有限元模型的有效构建方法进行了长时间的探索,然后构建出了解剖结构比较精细的正常足踝部有限元模型,分别在模型中模拟各种踝关节损伤及内固定术式,探讨其对足踝部生物力学的影响。目的：1.依据1名正常男性志愿者踝关节中立位下螺旋CT扫描的右侧踝关节面上20 cm以下的足踝部影像学资料,探讨利用Mimics、SolidWorks、ANSYS等软件构建出解剖结构精细的正常踝关节三维有限元模型的方法,并对其有效性进行验证,使其能反映正常足踝部的力学特性。2.根据所建立的正常足踝部的三维有限元模型,对其施加不同载荷,模拟人体静态中立位单足站立负重状态、踝关节受到内旋力、踝关节受到外旋力、踝关节受到内翻力及踝关节受到外翻力,探讨正常踝关节在不同活动方式中,主要关节周围各组织的应力、位移分布,以及关节活动度(range of motion, ROM)的变化等情况,同时对其结果的有效性进行验证。3.根据所建立的正常足踝部的三维有限元模型,模拟建立下胫腓联合损伤的模型,并在下胫腓联合损伤模型基础上分别建立踝关节平面上方的2.5 cm、5cm螺钉固定模型,探讨各模型在模拟人体中立位单足站立、踝关节内旋、及踝关节外旋加载中主要关节周围组织的应力、位移分布,以及关节活动度的变化等情况,同时对其结果的有效性进行验证；并通过与正常模型在这3种加载中的计算结果对照,探讨下胫腓联合损伤及踝关节平面上方2.5 cm、5 cm螺钉固定后对踝关节应力分布及稳定性的影响。4.根据所建立的正常足踝部的三维有限元模型,分别模拟建立后踝骨折累及1／5、1／4、1／3及1／2的踝关节面的损伤模型,探讨各模型在模拟人体中立位单足站立、踝关节内旋、及踝关节外旋加载中主要关节周围组织的应力、位移分布,以及关节活动度的变化等情况,同时对其结果的有效性进行验证；并通过与正常模型在这3种加载中的计算结果对照,探讨后踝骨折分别累及1／5、1／4、1／3及1／2的踝关节面后对踝关节应力分布及稳定性的影响。方法：1.数字化足踝部三维有限元模型的构建与验证：采用南方医科大学南方医院影像中心的LightSpeed 16排螺旋CT,扫描参数：电压120～140 KV,电流强度240～300 mA,螺距1.375～1.75,层厚7.5 mm,矩阵512×512,重建层厚0.625mm。对志愿者右足自踝关节上20 cm胫腓骨远端向下扫描至足底,扫描时右足保持中立位,以DICOM格式输入个人计算机的Mimics 10.01软件,经自动或手动阈值分割后三维重建出完整足踝部的28块骨骼及外围软组织的三维结构,再以点云输出并导入SolidWorks 2009,利用网格处理向导及曲面生成向导生成几何模型并重建装配体,然后分别导入两种有限元分析软件：(1)方法一为导入有限元分析软件ANSYS 12.0的Workbench模块建立完整的足踝部有限元模型。在踝关节的接触面两侧根据关节间隙分别建立了0.5 mm关节软骨,其他关节采用仅受压的三维杆单元模拟软骨,并建立了125条弹簧模拟韧带和小腿骨间膜,5条梁单元模拟跖筋膜,材料属性分别参照文献确定。然后参照Anderson等方法,对模型仅取胫骨及距骨所构成的简单踝关节模型加以测试。通过模拟人体单足站立状态下的力学传递,对模型胫骨下端的上截面施加600 N垂直载荷,对距骨予以约束。测量踝关节胫骨下关节面的接触压力、接触面积,并将结果与前者进行对照。(2)方法二为导入Simulation建立简化的踝关节有限元模型。根据研究需要,分别建立了包含胫骨、腓骨、距骨、跟骨及舟骨的5骨装配体,以及还包括骰骨及3块楔骨的9骨装配体的有限元模型。在模型中用“仅延伸”的弹簧模拟韧带连接,其中在踝关节5骨装配体中共建立了31条弹簧,而在9骨装配体中建立了42条弹簧模拟踝关节周围韧带及小腿骨间膜等连接装置。定义各组织的材料属性,并通过自动或手动方法在各关节间生成接触对,设置相应的边界条件分别在踝关节5骨装配体有限元模型上模拟人体单足站立及踝关节内、外旋运动加载,而在踝关节9骨装配体有限元模型上模拟踝关节内、外翻运动加载,然后设置好网格划分密度生成网格并设置算例属性进行运算。2.下胫腓联合损伤及固定术式的有限元分析：在前述方法二所建立的5骨装配体模型上分别将代表下胫腓联合前韧带、骨间韧带、后韧带及小腿远端8 cm的小腿骨间膜弹簧压缩(suppress),模拟建立下胫腓联合损伤的模型。然后在下胫腓联合损伤模型基础上分别在踝关节平面上方的2.5 cm、5 cm处平行于踝关节面并穿过腓骨、胫骨骨干的中央生成两个￠3.5mm孔,分别采用材料为316不锈钢的销钉连接两孔的圆柱面模拟建立两种固定方式的螺钉固定模型。然后设置好网格划分密度生成网格,对各种模型分别模拟人体中立位单足站立及踝关节内旋、外旋运动三种踝关节受力方式进行加载运算,并将计算结果与正常模型在这3种加载中的计算结果相对照。3.后踝骨折的有限元分析：在前述方法二所建立的5骨装配体模型上,在模型的胫骨下端踝关节面上按照文献报道的标准,分别切除相应大小的骨块,模拟后踝骨折累及1／5、1／4、1／3及1／2的踝关节面,同时删除累及损伤的韧带,建立4种后踝骨折损伤模型。然后设置好网格划分密度生成网格,对各种模型分别模拟人体中立位单足站立及踝关节内旋、外旋运动三种踝关节受力方式进行加载运算,并将计算结果与正常模型在这3种加载中的计算结果相对照。结果：1.在按方法一所建的完整的足踝部有限元模型中,将胫、距骨组成的踝关节测试结果与Anderson等实验结果对照后发现,二者在中立位踝关节垂直加载下应力主要分布于胫骨下关节面中部及前外侧,最大应力处于2.7 MPa～4.0 MPa之间,在分布区域和分布趋势、数值大小上基本一致,初步证明本模型是有效的。按方法二所建立的简化的两个踝关节三维有限元模型经测试可以模拟踝关节的不同受力情况,并可进行正确的加载运算,而且单元数、节点数、运算时间适中。各种加载后所产生的应力、位移的结果合理,比较贴近真实状况。将结果与Liacouras和Wayne在相同边界条件下所作的踝关节外旋模拟实验结果进行了对比,发现本模型在踝关节外旋加载中胫骨的旋转角度为3.85°,与其有限元模型计算的4.28°比较接近,另外本模型的主要关节的接触力大小也与其结果对应性较好,证明了本模型的有效性。2.对下胫腓联合损伤模型的研究发现,下胫腓联合损伤后会导致腓距关节的接触力降低,其他关节接触力增高,踝关节部分韧带抗阻力承受的载荷增大,胫骨、腓骨下端位移增大,胫骨的旋转角度(模型中代表踝关节旋转角度)有了明显增大,分别从正常的踝关节内旋时的7.110增加到了12.430,外旋时的3.85°增加到了5.46°,其中踝关节外旋加载结果显示与Liacouras和Wayne关于下胫腓联合损伤模型的研究结果,——胫骨的旋转角度在踝关节外旋模拟加载中从正常4.280增加到了损伤后为5.60°,对比非常接近,说明本损伤模型结果具有有效性。而后对两种螺钉固定模型模拟研究发现,各关节的接触力都比正常有所减少,胫骨、腓骨下端位移减少,胫骨的旋转角度明显缩小,腓骨旋转角度趋向于同胫骨接近,但这两种螺钉固定之间的差异不太明显。3.对累及损伤不同大小关节面的后踝骨折模型的研究发现,在中立位单足站立时,后踝骨折从损伤1／4踝关节面开始踝关节的最大应力明显增高,而在踝关节内旋加载中踝关节的最大应力是从损伤1／3踝关节面以上明显增高；各骨折模型在中立位单足站立及踝关节内旋加载中均表现为腓距关节的接触力降低,胫距(踝)关节接触力增高；另外,胫骨的旋转角度在踝关节内旋加载时随着后踝骨折损伤关节面的增大角度呈递减趋势,而在踝关节外旋加载中,其角度又随着损伤程度加重而表现为递增现象。结论：1.本研究利用Mimics、SolidWorks、ANSYS等软件依据螺旋CT扫描的二维图像构建出了解剖结构精细的正常人体足踝部三维有限元模型,相对客观地反映了足踝部的基本解剖结构和力学特性,模型经验证有效,并可在模型上模拟足踝部不同受力方式进行力学分析。2.通过对下胫腓联合损伤及两种螺钉内固定的模型结果分析显示,下胫腓联合损伤会导致胫、腓骨下端位移增加,使踝关节在内旋与外旋活动中关节活动度增加,从而引起关节不稳定；而在使用两种位置的下胫腓联合螺钉固定后,均表现为关节活动中各骨位移减少,关节活动受限,关节接触力减小,但两种位置的螺钉之间对比差异性不大,因此,其并非属于符合生理的固定,只适宜短期使用。3.通过对累及损伤不同大小踝关节面的后踝骨折模型的结果分析表明,尽管小的后踝骨折(损伤1／4以下关节面)在本研究中显示没有明显增加踝关节的最大应力,临床中可考虑保守治疗,但各种后踝骨折都会引起关节接触力及关节活动度发生明显改变。更多还原

【Abstract】 Background:With the modernization process, there having an unprecedented development in transportation industry and construction business, accompany with the improvement of people’s health concept, foot and ankle diseases have been increasing and becoming more complex each year, which makes the orthopedics face new challenges clinically and urgently needs to be explored in depth. Foot is the base of human body’s weight-bearing, and direct contact with the ground. In addition to having walking function, it’s also able to absorb the movement oscillation, cushion ground reaction force to protect vital organs and organizations of the body from being harmed. Meanwhile, it can make the body’s center of gravity stable. The ankle joint is the junction of human body in contact with the ground. Walking, jumping, running, and climbing all need the involvement of the ankle, even riding a bicycle or driving a car it also being indispensable to coordinate the motion. It can be said that there is no movement in daily life whithout the involvement of the ankle, so it is one of the the most susceptible joints to be hurt. The ankle joint has complex mechanical mechanism. A variety of injuries are likely to break the balance of its surrounding structures and cause mechanical instability, which can finally induce traumatic arthritis. Furthermore, in adition to leading to pain, deformity and activity disorders, structural abnormalities of foot and ankle can further affect mechanical functions of lower extremities, pelvis, spine and other body part. Therfore, it has important clinical significance to study on them.Foot and ankle biomechanics research is an important section of human biomechanics study. Many researchers pointed out that the biomechanical factors play an important role in exploring the etiology, pathogenesis, treatment and prevention of the disease of foot and ankle. In the past, studies on foot and ankle mainly used traditional experimental means of biomechanical tests. But the disadvantage of this method is the complexity of experimental means, and when carried out a variety of operating conditions, the experiment is often costly, time-consuming and low efficiency. Moreover, because it can not directly test in humans, it is difficult to accurately reflect the stress distribution after application different load. In order to solve this problem, in recent years, with the improvement of digital technology, more and more researchers use numerical analysis methods, namely, based on the theory of traditional mechanical analysis, numerical analysis methods such as finite element method (finite element method, FEM) and so on, for linear and nonlinear stress and deformation analysis. It can be simulated using a variety of foot and ankle disorders by FEM, making biomechanical studies on bones’ complex geometric structures, boundary conditions and material nonuniformity of the foot and ankle issues have possible solutions.FEM has powerful capabilities of modeling and can simulate complex geometric structures, material parameters and different force in the dynamic and static state, which has increasingly been applied to the human body biomechanics. Because of the complexity of bones, joints, ligaments, tendons and other structures of foot and ankle, too much work in finite element modeling as well as the lack of detailed parameters of the various organizations, it is more difficult to simulate than other large joints by FEM. Existing finite element models of foot and ankle, most of them have not constructed anatomic structure of the ankle joint perfectly, mainly for the research related to plantar pressure. However, there is clinically growing recognition to the effects of ankle injuries on human body functions, urgent needs for more in-depth study on them.Based on previous studies, we have taken a long time to explore the effective methods of modeling, then established finite element models of nomal human foot and ankle with fine anatomical structures, through simulating many kinds of ankle joint injuries and internal fixation pattern in the models, to explore their biomechanics effect on foot and ankle.Obiective1. According to spiral CT scan images of a normal male volunteers’ right foot and ankle from the plane 20 cm above the ankle down to the planta with ankle joint neutral position, to explore how to establish finite element model of nomal human foot and ankle with fine anatomical structures with Mimics, SolidWorks and ANSYS softwares. Moreover, its validity should be verified, so that it can reflect the mechanical characteristics of the normal foot and ankle.2. According to the three-dimensional finite element model of normal foot and ankle established, apply different loads to simulate the static weight-bearing state of human body in neutral position with one foot standing, and states of internal rotation of ankle, external rotation of ankle, ankle inversion, and ankle eversion, so as to explore the stress, displacement distributions of surrounding tissue, and changes of joint range of motion in main joints during various forms of normal ankle motion. At the same time, the validity of their results to be verified.3. According to the three-dimensional finite element model of normal foot and ankle established, to simulate the establishment of distal tibiofibular syndesmosis injury model, and based on distal tibiofibular joint injury models, to establish two kind of screw fixation models, in which the screw inserted 2.5 cm,5 cm above of the ankle respectively, then explore the stress, displacement distributions of surrounding tissue, and changes of joint range of motion in main joints after application of neutral position with single foot standing, internal rotation and external rotation loads. At the same time, the validity of their results to be verified. Moreover, by comparison with the results of normal model, to explore the effect of distal tibiofibular syndesmosis injury and screw fixation inserted 2.5 cm,5 cm above of the ankle respectively on stress distribution and stability of ankle.4. According to the three-dimensional finite element model of normal foot and ankle established, to simulate the establishment of posterior malleolus fracture involving 1/5,1/4,1/3 and 1/2 of the ankle joint surface injury respectively, then explore the stress, displacement distributions of surrounding tissue, and changes of joint range of motion in main joints after application of neutral position with single foot standing, internal rotation and external rotation loads. At the same time, the validity of their results to be verified. Moreover, by comparison with the results of normal model, to explore the effect of posterior malleolus fracture involving 1/5,1/4, 1/3 and 1/2 of the ankle joint surface injury respectively on stress distribution and stability of ankle.Methods1. Construction and verification of digital three-dimensional finite element model of foot and ankle:Adopting LightSpeed 16-slice spiral CT of the Imaging Center, Nanfang Hospital affiliated to Southern Medical University (scanning parameters:120～140 KV,240～300 mA, pitch of 1.375～1.75, the layer thickness 7.5 mm, matrix 512×512, reconstruction slice thickness 0.625 mm), scaned the volunteer’s right foot from distal tibia and fibula 20 cm above the ankle down to the planta, with the right foot remaining neutral position. Then imported the scanned data of DICOM format into Mimics 10.01 software, by threshold segmentation automatically or manually, to reconstruct the three-dimensional structure of a complete foot and ankle composed of 28 bones and surrounding soft tissue. Finally, exported the data with point cloud format and reimported into SolidWorks 2009, using the guide of grid processing and surface generation to form geometric models and reconfigure them, then import the data into two kinds of finite element analysis softwares:(1) Method 1:Imported into Workbench module of finite element analysis software ANSYS 12.0 to establish a complete finite element model of foot and ankle. To construct 0.5 mm of articular cartilage on both sides of contact surface according to the joint space, while to use three-dimensional rod elements which were compressed only to simulate other joints cartilage, and to establish 125 springs to simulate ligaments and crural interosseous membrane, five beam elements to simulate plantar fascia.The material properties were determined with reference to documents. Then refered to Anderson’s method, took the tibia and talus only as a simple model of the ankle joint for test. The normal standing status of ankle joint was simulated by application a vertical load of 600 N on the upper section of the lower tibia while the talus constrained. To measure contact pressure and contact area of inferior articular surface of the tibia, and to compare the results with the former. (2) Method 2: Imported the data into Simulation module of Solidworks to establish a simplified finite element model of the ankle. According to the research needs, to establish a 5-bones assembly finite element model containing the tibia, fibula, talus, calcaneus, and navicular, and a 9-bones assembly finite element model also including the cuboid bone and the three cuneiform bones in addition to the 5 bones mentioned above. In the models, to use tension-only springs to simulate ligaments connection. The 5-bones assembly contained 31 springs, while the 9-bones assembly was established 42 springs to simulate connected structures such as ligaments around the ankle and crural interosseous membrane. To Define the material properties of each tissue, to generate contact pair between each joints automatically or manually, and to set the corresponding boundary conditions. In the 5-bone assembly finite element model of ankle, to simulate the state of human body with one foot standing and the states of internal and external rotation of ankle, while in the 9-bones finite element model of ankle, to simulate the states of ankle inversion and ankle eversion. Then, regulated the mesh density to generate mesh, and set the simulation examples attribute for solution.2. Finite element analysis of distal tibiofibular syndesmosis injury and fixation pattern:based on the 5-bone assembly model constructed with aforementioned method 2, to establish distal tibiofibular syndesmosis injury model by suppressing the springs represented the anterior and posterior tibiofibular ligaments(ATIFL, PTIFL), the interosseous ligament(IL), along with 8 cm of the distal interosseous membrane closest to the tibiotalar joint. Then based on the distal tibiofibular syndesmosis injury model, at the plane 2.5 cm,5 cm above the ankle joint generated two￠3.5mm holes respectively, which parallel to ankle joint surface and through the center of fibula and tibia diaphysis. The cylindrical surfaces of the holes were connected with 316 stainless steel pin to simulate the establishment of two kind of screw fixation models. After that, to regulate the mesh density to generate mesh, three loads such as neutral position with single foot standing, internal rotation and external rotation of ankle were simulated for solution. Finally, to compare the results with the normal model.3. Finite element analysis of posterior malleolus fracture:based on the 5-bone assembly model constructed with aforementioned method two, to establish posterior malleolus fracture model involving 1/5,1/4,1/3 and 1/2 of the ankle joint surface injury by resection the corresponding size of fragment in accordance with the standards reported in the literature, respectively. At the same time, involving ligaments were also deleted. After that, to regulate the mesh density to generate mesh, three loads such as neutral position with single foot standing, internal rotation and external rotation of ankle were simulated for solution. Finally, to compare the results with the normal model.Results1. In the complete finite element model of foot and ankle constructed by method 1, the test results of the ankle composed of tibia and talus were taken to contrast with Anderson’s study. It showed that the stress on ankle joint articular surface of the tibia mainly distributed at the central and anterolateral aspects in both of them when ankle joint was in neutral position under the vertical loading, with the maximum stress between 2.7 MPa and 4.0 MPa, substantial agreement on distribution area, distribution trend, and numerical value. It was preliminarily verified this model valid. The test results of two simplified three-dimensional finite element models of the ankle established by method 2 showed that they could simulate various load situations of ankle for solution, with moderate number of element, node and suitable amount of computing time. The stress and displacement results in each loading were reasonable and close to the actual situation. Compared with Liacouras and Wayne’s model under the same boundary conditions, it showed that the rotation angle of tibia was 3.85°in our model whereas their result was 4.28°after application of ankle external rotation load, comparatively close between the two values. In adition, the contact forces of major joints were also consistent with theirs. All confirmed the validity of our models.2. The studies of distal tibiofibular syndesmosis injury model demonstrated that distal tibiofibular syndesmosis injury would reduce contact forces between the talus and fibula whereas raise other joints’ contact forces, make some ligaments of ankle withstand more load resisting movement, increase the magnitude of displacement at the lower extreme of tibia and fibula, augment the rotation angle of tibia(represented the rotation of ankle in the model) from normal intactness 7.11°to injuried 12.43°after application of internal rotation load and from 3.85°to 5.46°after application of external rotation load respectively, among which, the rotation angle of ankle external rotation loading was very close to Liacouras and Wayne’s result from 4.28°to 5.60°. It was further confirmed this injury model valid. While the succeeding two kinds of modes fixed with screw showed distal tibiofibular syndesmosis injury fixed with screw would reduce contact forces more or less in all joints, decrease the magnitude of displacement at the lower extreme of tibia and fibula, distinctly diminish the rotation angle of tibia, make the rotation angle of fibular close to tibia. However, there were less obvious differences between the two kinds of ways of screw fixation.3. The studies of posterior malleolar fracture models involving different size of ankle joint surface injuries demonstrated that the maximum stress of ankle joint surface increased obviously from posterior malleolar fracture involving 1/4 of ankle joint surface during application of neutral position with single foot standing load, whereas the maximum stress increased from posterior malleolar fracture involving 1/3 of ankle joint surface during application of ankle internal rotation load; each model of posterior malleolar fracture all demonstrated contact forces between the talus and fibula reduced whereas contact forces between the tibia and talus(ankle jiont) raised; In addition, the rotation angle of tibia was decreased with posterior malleolar fracture models involving ankle joint surface increased during application of ankle internal rotation load, whereas the rotation angle was increased with posterior malleolar fracture models involving ankle joint surface increased during application of ankle external rotation load.Conclusions1. This study has successfully constructed three-dimensional finite element models of normal human foot and ankle with fine anatomical structures by means of Mimics, SolidWorks, ANSYS softwares in accordance with two-dimensional images scaned by spiral CT, and reflected the basic anatomic structures and mechanical properties of the human foot and ankle.The models were proved to be valid and could simulate various load situations of ankle for mechanical analysis.2. The analytic results of distal tibiofibular syndesmosis injury model and two kinds of fixation modes with screw showed that distal tibiofibular syndesmosis injury would increase the magnitude of displacement at the lower extreme of tibia and fibula, make joint ROM increase during internal rotation and external rotation of ankle, which led to joint instability, whereas, after using the distal tibiofibular screw fixed in two different positions, the results of them both represented the magnitude of bones displacement decreased, the motion of joints limited, and contact forces of joints reduced. However, the difference between the two kinds of ways of screw fixation inserted was not significant. Therefore, the fixation with screw didn’t belong to physiological, and its present role was restricted to short term applications.3. The analytic results of posterior malleolar fracture models involving different size of ankle joint surface showed that dispite the posterior malleolar fracture with small fragment (injury less than 1/4 of the joint surface) seemed not to increase the maximum stress of ankle joint, conservative treatment could be considered clinically. However, all kinds of fractures could cause contact forces and ROM of joint changed significantly.更多还原