【作者】 李丹；

【导师】 谷贵山；

【作者基本信息】 吉林大学 ， 外科学， 2011， 博士

【摘要】 目前有限元分析的发展已经从研究对象的生理阶段进入到病理化阶段。有限元研究的病理化阶段的研究由于病例数据来源匮乏,存在以下三个缺陷,即以局部代替整体(单一椎体代替整个腰椎),以个体代替群体(研究只用1例),以生理代替病理(正常脊柱代入异常数值)。本研究引入全新的研究对象选取方法,克服有限元分析病理化阶段的上述三个缺陷,选取整个腰椎而非单一椎体,选取多组病例而非单一病例,选取典型病例而非正常脊柱,探讨以真实病理状态的腰椎作为研究对象揭示病理状态的真实应力分布,从力学角度深入认识脊柱疾病的病理状态产生的诱因及后果。材料和方法：(1)如何科学的选择典型病例进行病理生物力学有限元分析,就成为我们面临的主要问题。本研究采用了PACS系统和电子病历系统(Electronic Medical Record, EMR),对9670例腰椎CT检查病例的选取及筛查。筛查的入组标准为骨质疏松脆性骨折(从站立高度摔倒发生的骨折)即老年骨折组、年轻对照组及老年非骨折对照组。排除标准如下：其中排除暴力骨折1301例,排除骨代谢相关疾病(如：甲状旁腺机能亢进、肾性骨病等),排除患者有转移性恶性肿瘤病史、原发骨肿瘤病史,排除患者有肝硬化,糖尿病I型,及糖尿病II型的病人。根据入组及排除标准,共751例患者进入本研究第一步筛查有统计学意义的骨折相关指标。利用MPR, MIP, VR技术,每个椎体观察指标为10项,其中变量20个,整个腰椎的形态学变量为104个。获得了老年骨折组、老年非骨折组及正常腰椎的形态学数据。对上述数据进行Logistic回归分析筛选脆性骨折相关变量指标。(2)通过SYNGO软件对原始腰椎DICOM数据进行,组织标记、三维切割、三维定位归位、平面重建之后,和MIMICS进行DICOM数据交接,整个腰椎的模型建立完成。对椎体、椎间盘模型赋予材料属性,用ABAQUS进行FEA分析。(3)不同研究组腰椎的有限元分析：以统计学数据为依据,选取各项指标位于中位数的典型病例。年龄差异组,这一组包括四个模型分别为wj, wj-f, zj, dhy。以观察椎间盘退变(材料属性改变),椎间盘间隙改变(形态属性改变)在不同年龄组对腰椎生物力学的影响。(4)老年女性腰椎椎间盘退变和腰椎骨折风险的FEA分析,以统计学数据为依据,选取各项指标位于中位数的典型病例。为观察椎间隙差异对生物力学的影响,这一组包括五个模型分别为wj, wj-f, dhy, nfx, ylq。选择的目的是尽量的保证CT值相同,赋予间盘相同的退变材料属性,没有其它因素的干扰,模型之间只有椎间隙高度(形态属性)的不同,讨论间盘间隙狭窄对腰椎节段生物力学的影响。此外还要观察胸12骨折是否加大骨折风险,导致骨折的级联效应。结果：在形态学分析中我们得知：在腰1-骶1椎间隙测量中与椎间隙间隙0级相比,狭窄程度1级到3级,腰1到腰5腰椎骨折风险率OR值分别从1.18上升到38.11,1.56上升到12.8,2.67上升到9.9,1.94上升到12.68,1.39上升到7.95。尤其是L1-2间盘间隙的减小使骨折风险上升迅猛。L1-2椎间隙后缘高度小于0.32cm时,是后缘高度大于0.48cm骨折风险的18.54倍。不同年龄组腰椎的有限元分析中：年龄差异组的直立,前屈,后伸工况下椎体,椎间盘,上下终板的FEA分析结果显示：不但椎间盘退变的材料属性影响腰椎的应力分布,不同年龄组腰椎、椎体与间盘退变的形态学改变也影响应力分布。老年女性腰椎椎间盘退变和腰椎骨折风险的FEA分析中：狭窄差异组的直立,前屈,后伸工况椎体,椎间盘,上下终板的FEA分析结果显示：椎间隙狭窄的加重,使应力更加集中。已经有一个椎体骨折的病人,发生级联骨折的机会加大。结论：椎间隙间隙的降低与腰椎骨折风险呈显著正相关。FEA生物力学分析证实,除了材料属性外,无论是椎间盘的形态属性还是椎体的形态属性,乃至脊柱整体形态的结构属性对腰椎生物力学应力分布的影响都是至关重要的。提出病理的形态结构与病理的应力分布交互作用的新观点,即病理的结构形态产生病理的应力分布,病理的应力分布反过来加重病理的形态结构改变。提出腰椎-椎体-间盘-椎体-腰椎相互作用影响腰椎生物力学性能增加骨折风险的新观点。本研究的创新点：(1)有限元生物力学分析的病理化(生物力学分析研究对象选取的创新)：目前有限元分析的发展已经从研究对象的生理阶段进入到病理化阶段。有限元研究的病理化阶段存在的主要问题：由于病例数据来源匮乏,存在以下三个缺陷,即以局部代替整体(单(?)体代替整个腰椎),以个体代替群体(研究只用1例),以生理代替病理(正常脊柱代入异常数值)。本研究引入全新的研究对象选取方法,克服上述三个缺陷,选取整个腰椎而非单一椎体,选取多组病例而非单一病例,选取典型病例而非正常脊柱,真正做到以病理状态的腰椎作为研究对象揭示病理状态的真实应力分布,从力学角度深入认识脊柱疾病的病理形态结构产生病理应力及病理应力反过来引起的进一步的结构形态的退变。本研究完成了生物力学分析研究对象选取的创新,这使得我们的有限元分析可以把真正病理状态的整个腰椎作为研究对象,把有限元生物力学分析真正引入病理化阶段。(2)病例选择的典型化科学化(科学选取典型病例的创新)：通过大规模病例的选取及筛查,选取10大类104项指标综合反映腰椎的病例形态特征,以统计学数据为依据,选取各项指标位于中位数的典型病例,大大提升了本研究的数据收集整理的科学性、可靠性和可行性。本研究完成了病例选取科学性的创新,使得有限元分析生物力学分析能够基于典型病例的典型形态学及材料学特征,反映出典型病理性应力分布特征。为腰椎疾病的发生发展和转归提供科学可靠的依据。实现了病例选择的典型化。(3)模型建立的自动化、标准化、准确化(模型建立方法的创新)：由于有限元分析目前广泛使用的MIMICS软件的功能有限,有限元分析建模的速度和准确性成为研究的主要障碍,以至于多节段重建和大样本量分析的可行性受到限制。现阶段有限元分析基于MIMICS软件,建模的过程需要将每个扫描层面手工分辨组织边界切割处理,在目前CT扫描发展到动辄成百上千层的信息,使建模的工作不仅庞杂缓慢,还因为手工切割的主观性误差丢失重要的影像信息。本研究开创性的应用SIEMENS后处理SYNGO3D软件,进行原始数据软件自动识别标记分割预处理,并成功与MIMICS软件进行数据交接。这种自动高分辨率组织识别和三维切割大大的提高了建模工作效率,不但使FEA模型建立过程缩短了10倍的时间,并且能够获得更精准的椎体模型信息。完成了有限元分析模型建立方法的创新。实现了模型建立的自动化、标准化、准确化。使得腰椎多节段、大样本量有限元分析成为可能。(4)有限元生物力学分析结果的创新：本研究提出了病理状态生物力学有限元研究的新方法和新结论,FEA生物力学分析证实,除了材料属性外,无论是椎间盘的形态属性还是椎体的形态属性,乃至脊柱整体形态的结构属性对腰椎生物力学应力分布的影响都是至关重要的。提出病理的形态结构与病理的应力分布交互作用的新观点,即病理的结构形态产生病理的应力分布,病理的应力分布反过来加重病理的形态结构改变。提出腰椎-椎体-间盘-椎体-腰椎相互作用影响腰椎生物力学性能增加骨折风险的新观点。上述创新点在国内外同类研究中未见报道。更多还原

【Abstract】 Currently, the development of finite element analysis (FEA) has entered into the era in which the objective of study is shifting from physiological status to pathological status. At the beginning of pathological era of FEA, with the scarcity of the source of pathologic data, there are three major limitations in FEA study:analysis of one segment instead of the whole lumbar spine, analysis of only one case instead of the population with the same disease, analysis of pseudo-pathological status by simply assign abnormal material properties into an entirely morphologically normal lumbar spine.To overcome the limitations of the pathological era of FEA listed above, our study introduced a new method of study object selection, which included the whole lumbar spine instead of one segment; include multiple cases instead of solely one case, included typical pathologic cases instead of pseudo-pathological cases with normal morphology. The FEA study based on the genuine pathologic lumbar spine will reveal the actual stress distribution of pathologic status which is far more close to the reality. The pathological FEA study would provide us with a better understanding of pathologic status of spinal disease and corresp-onding biomechanical remote cause and subsequent prognosis.Materials and methods:(1) Case selection:The inclusion criterion of case selection is essential for the typical characteristics of morphological properties of the lumbar spine and subsequent accurate analysis of FEA study. During the past three years (from Jan.2008 to Dec.2010) 9670 lumbar CT examinations were taken and recorded in the PACS(picture archiving and communication system) and EMR (Electronic Medical Record) system in our hospital. The inclusion criteria for screening are fragile fractures (fractures happen from falling down at standing height) for senile fracture group, and young control group and senile non-fracture control group. The exclusion criteria are violent lumbar fractures (1301 cases), bone metabolism related diseases (such as hyperparathyroidism, nephritic bone disease, and so on), metastatic malignant tumors, primary spine tumors, hepatic cirrhosis, typeⅠdiabetes, typeⅡdiabetes. With the inclusion and exclusion criteria,751 cases were enrolled in our study for the first step of screening for fracture-related indices with significant statistical correlation. The screening and selection of the typical cases of lumbar spine was based on the three-dimensional lumber morphological data evaluated through the technique of MPR (multi-planer reconstruction), MIP (maximum intensity projection), and VR (volume rendering). There were 20 variants for each vertebra. The overall morphological variants of the whole lumbar spine were 104. The morphological variants of senile fracture group, senile non-fracture group and young control group were obtained and screened for fragile fracture related indices through logistic regression analysis.(2) The raw DICOM data of the lumbar spine underwent the processing of tissue mark, three dimensional cutting, three dimensional home-position, multi-planer reconstruction with the software of SYNGO on the Workstation of SIEMENS. After the DICOM data transmission from SYNGO to MIMICS, the models of lumbar spine were set up. The FEA analyses were carried out by the software of ABAQUS with the assignment of material properties for both vertebrae and intervertebral discs.(3) The FEA analysis of lumbar spine of different groups:The typical cases of each group were selected according to the median value of the correlated indices. To observe the effect of changes of material properties of the lumbar disc (disc degeneration) and the effect of changes of morphological properties (disc space narrowing) on the biomechanics of the lumbar spine in different groups of degeneration, four typical cases of models were established as wj, wj-f, zj, and dhy, respectively.(4) The FEA analysis of the correlation between the fracture risk and the degeneration of lumbar disc in senile females:The typical cases of each group were selected according to the median value of the correlated indices. To observe the effect of degree of intervertebral disc degeneration on the changes of biomechanics of lumbar spine, five typical cases of models were established as wj, wj-f, dhy, nfx, and ylq, respectively. The objective of this study was to observe the degree of disc space narrowing on the biomechanical changes of the lumbar spine without the interference of other factors. Therefore, the monofactorial analysis was guaranteed by the case selection with the same CT value, assignment of the same elastic modulus for the degenerated disc, with the sole variant of intervertebral disc height. Moreover, we should observe whether there was the cascade effect of higher fracture risk correlated with the fractures already existed on the vertebra T12.ResultsThe evaluation of the disc space shows that the risk of vertebral fracture increased with the severity of disc space narrowing. Based on the odds ratio (OR) for grade 0 disc space, the changes of OR value of grade 1 disc space and grade 3 disc space for L1 to L5 are listed below:from 1.18 to 39.11 in L1, from 1.56 to 12.8 in L2, from 2.67 to 9.9 in L3, from 1.94 to 12.68 in L4, from 1.39 to 7.95 in L5, respectively. The disc space of L1-2 showed the most evident negative correlation with vertebral fracture risk. In comparison with the cases of disc distance of posterior height (DPH) greater than 0.48cm in L1-2 disc, the cases of DPH less than 0.32cm in L1-2 disc showed an increase of 18.54 times of vertebral fracture risk.In the FEA study of different age groups, the stress distribution status of vertebral bodies, intervertebral discs and endplates were observed under different loading condition of neutral standing, flexion and extension. The FEA results of vertebral body, intervertebral disc and endplate showed that not only the degenerative changes of material property of the intervertebral disc had an effect on the stress distribution of the lumbar spine; morphological changes of vertebral body and intervertebral disc also played an important role in the stress distribution in lumbar spine between different age groups.In the FEA study of correlation between lumbar disc degeneration and vertebral fracture risk in senile groups, the stress distribution status of vertebral bodies, intervertebral discs and endplates were observed under different loading condition of neutral standing, flexion and extension. The FEA results showed that the more disc space height reduction, the more stress concentration. The cascade fracture risk increases in case of an existed vertebral fracture.Conclusions:Disc space narrowing showed an obvious positive correlation with vertebral fracture risk. Based on FEA biomechanical analysis, we present a further new point of view that not only the material properties, but also the morphological properties of both intervertebral discs and vertebrae, and even the macromorphology of the whole lumbar spine will play an important role in the biomechanical stress distribution of lumbar spine. We presented an until-now unrecognized new view on the interactive relations between pathological stress distribution and pathological structure and morphology. That is to say, pathological structure and morphology generates the pathological stress distribution, and the pathological stress distribution triggers further morphological and structural degeneration. The interactive relation of the spine (spine-vertebra-disc-vertebra-spine) increases the risk of osteoporotic vertebral fractures.The innovation points of our study:(1) The innovation of objective selection in FEA study:We carried out FEA study on the genuine pathologic cases to reveal the actual stress distribution of pathologic status which is far more close to the reality. Currently, the development of finite element analysis (FEA) has entered into the era in which the objective of study is shifting from physiological status to pathological status. At the beginning of pathological era of FEA, with the scarcity of the source of pathologic data, there are three major limitations in FEA study:analysis of one segment instead of the whole lumbar spine, analysis of only one case instead of the population with the same disease, analysis of pseudo-pathological status by simply assign abnormal material properties into an entirely morphologically normal lumbar spine. To overcome the limitations of the pathological era of FEA listed above, our study introduced a new method of study object selection, which included the whole lumbar spine instead of one segment; include multiple cases instead of solely one case, included typical pathologic cases instead of pseudo-pathological cases with normal morphology. The FEA study based on the genuine pathologic lumbar spine will reveal the actual stress distribution of pathologic status which is far more close to the reality. The pathological FEA study would provide us with a better understanding of pathologic stress distribution of lumbar spine and corresponding effect of pathologic stress distribution on the further degeneration of intervertebral discs and vertebrae. The innovation of objective selection in our FEA study enrolled the whole pathologic lumbar spine as objectives, which fulfill the genuine pathologic FEA study.(2) The innovation of scientific selection of typical cases:The typical case selection was based on the screening of 10 categories of 104 morphological indices, which could reflect the overall morphological properties of the lumbar spine. Fragile fracture related indices were determined through logistic regression analysis. The selections of typical cases in each group were determined according to the median value of the correlated indices. The scientific selection of typical cases was fulfilled through screening in a large sample with statistical analysis. The cases with indices at median values provided us with typical morphological and material properties, which could reflect typical biomechanical changes in FEA study. The screening and statistical analysis in case selection guaranteed us with a scientific, reliable and feasible source of data for FEA analysis. The typical biomechanical changes of typical cases would provide us with a better understanding of typical pathological stress distribution characteristics and the corresponding causes, development, and prognosis of the lumbar diseases.(3) The innovation of FEA model construction (automatic, standardized, and accurate construction of FEA model):Currently, as the most widely-used software in FEA model construction, MIMICS is limited in the functions of three- dimensional processing, which hindered the speed and accuracy of FEA model construction. Therefore, the feasibility of multi-segment analysis and large sample analysis is restricted. The model construction on the software of MIMICS is conducted manually. The margin between bone and soft tissue is discerned and cut manually slice by slice, which is not only time-consuming but also not accurate in tissue margin determination. The subjective error in margin determination will lead to loss of important data of image. The modern CT data are consisted of hundreds of slices, which made the work of model construction tedious with low accuracy. We innovatively introduced the method of three-dimensional image processing procedure on SYNGO 3D software of SIEMENS workstation before DICOM data transmission to the MIMICS. The raw DICOM data of the lumbar spine underwent the processing of automatic tissue mark, three dimensional cutting, three dimensional home-position, multi-planer reconstruction with the software of SYNGO on SIEMENS workstation. After the successful DICOM data transmission from SYNGO to MIMICS, the models of lumbar spine were set up. The technique of high-resolution automatic tissue mark and three dimensional cutting greatly improved the efficiency and accuracy of model construction in the FEA study. The time of model construction was ten times less than before. The automatic, standardized, and accurate construction of FEA model made it feasible for the multi-segment analysis and large sample analysis of lumbar spine.(4) The innovation of the results of FEA analysis:Based on FEA biomechanical analysis, we present a further new point of view that not only the material properties, but also the morphological properties of both intervertebral discs and vertebrae, and even the macromorphology of the whole lumbar spine will play an important role in the biomechanical stress distribution of lumbar spine. We present an until-now unrecognized new view on the interactive relations between pathological stress distribution and pathological structure and morphology. That is to say, pathological structure and morphology generates the pathological stress distribution, and the pathological stress distribution triggers further morphological and structural degeneration. The interactive relation of the spine (spine-vertebra-disc-vertebra-spine) increases the risk of osteoporotic vertebral fractures.The points of innovations listed above have not been reported up to date in the published literature both domestic and abroad.更多还原