【作者】 杨长伟；

【导师】 李明； 张春才； 张少成； 何大为； 傅强； 贺石生；

【作者基本信息】 第二军医大学 ， 外科学， 2009， 博士

【摘要】 一、研究背景人工椎间盘置换手术是最近几年出现的治疗腰椎退变性疾病的新方法,与传统的腰椎融合手术相比,其保留了病变节段的活动,更加符合生理状况,因而能够减轻融合手术后常见的邻近节段退变的并发症发生,具有很好的应用前景。但人工椎间盘置换手术也存在一些问题:有报道称术后小关节退变加速,并且也还存在邻近节段的退变问题;此外也存在与其他一些手术类似的并发症:如减压不彻底、神经功能损害等。而且,人工椎间盘置换手术患者相对比较年轻,术后需要长时间的随访,在这一较长的随访时间段,也可能出现其他脊柱相关的疾病。磁共振成像由于其优良的软组织对比性能,是脊柱外科诊断随访的有用工具,但是对带有金属植入物的患者进行磁共振检查时要考虑两个问题:一是成像安全性的问题,包括铁磁性材料在磁共振仪器强磁场的作用下发生作用而移位,及在射频作用下的产热问题。二是金属植入物在磁共振图像上的伪影问题,这会降低图像的诊断性能。但由于人工椎间关节置换手术越来越广泛,MRI对其术后随访的意义非常重大。关于金属植入物磁共振成像的安全性和相容性的问题,国际上正在建立相关检测标准,美国材料测试学会(American Society for Testing and Materials ,ASTM)研究制定了检测医疗器械在MR环境中安全性和有效性的适用方法和指导原则,到目前为止,已经形成了关于植入材料在MR环境中安全性和相容性的五个相关标准。关于MRI金属伪影,虽然难以完全避免,但也可通过一些方法减少其影响,如选择合适材料、改变患者体位或编码方向,采用短回波时间,低磁场强度,宽读出频带和小体素可减轻伪影,避免伪影较大的序列如快速梯度回波序列(gradient recalled echo,GRE),采用快速自旋回波(fast spin echo,FSE同TSE)和自旋回波(spin echo,SE)等序列以及新成像技术如视角倾斜可减少MRI金属伪影。反转回复序列(inversion recovery,IR)比频率选择脂肪饱和序列(frequency selective fat saturation)伪影小。然而大多数能减少金属伪影的措施也同样降低了磁共振图像的质量,特别是信噪比(signal-to-noise ratio,SNR)。目前经FDA批准的临床使用较多的腰人工椎间盘有:DePuy Spine公司的SB CHARITé和Synthes Spine公司的PRODISC?-L假体。两种假体均由成分为钴铬钼合金的两块盖板及聚合材料的中间体组成。有关两种假体的磁共振成像安全性、伪影表现及其对邻近结构的影响目前仍未见详细报道。本文将设计对照实验阐明其安全性有关的数据和伪影表现及其对邻近结构的影响,并提出优化的磁共振扫描方案,从而为腰椎间盘置换术后磁共振检查提供理论依据。二、研究目的1.通过对两种人工腰椎间盘在1.5TMRI磁场中受力以及在人体标本中MRI扫描后产热数据的测定,评价两种人工椎间盘在1.5T磁共振检查中的安全性。2.通过自行设计的伪影评价水模实现对两种人工椎间盘金属伪影的定量测量,并比较不同序列及参数条件对伪影大小的影响。3.通过标本模型,验证优化参数对人工椎间盘植入术后磁共振成像质量的影响。三、研究内容第一部分CHARITé与PRODISC?-L腰人工椎间盘磁共振成像安全性的研究根据ASTM F2052-06el推荐的磁共振主磁场受力测试方法,分别将两种人工椎间盘金属盖板用细线悬吊,用角度测量仪测出器械在磁共振入口处的最大偏移角度;采用新鲜人尸体标本,按照器械公司推荐的植入方法在L5/S1节段植入人工椎间盘,采用热敏温度计分别测量磁共振检查前和检查后植入物邻近组织的温度(快速自旋回波序列,极端射频波,作用30分钟)。第二部分人工椎间盘磁共振伪影分析水模的设计及两种人工椎间盘在不同扫描序列及参数下的伪影分析根据ASTM推荐的伪影测试方法,采用自行设计的水模来测试伪影,水模为有机玻璃材料组成的长方体,内盛1%CuSO4溶液,用细线将椎间盘假体组合悬吊固定于水模中央。在植入与不植入人工椎间盘的情况下,分别采用SE、TSE、GRE、短T1反转回复(Short T1 Inversion Recovery,STIR)等序列及调整参数(视野、相位编码方向、回波时间、饱和带、带宽、回波链长度等)进行检查,得到植入物伪影图像,对伪影形态进行描述,并根据ASTM推荐的伪影测试方法,在OSIRIS影像处理软件上测量伪影的最大直径,比较不同序列下伪影大小,并探讨不同参数对伪影大小的影响。第三部分优化参数在人工椎间盘植入术后磁共振成像中的应用在人新鲜尸体标本L5/S1椎间植入人工椎间盘,在标本上分别采用经水模实验优化的TSE序列参数和非优化参数成像,分别取矢状位和横切位两次检查相同层面典型图像存为JPEG格式文件,编号,寄给6位脊柱外科或影像科专家。采用Jarvik评分法分别对几个重要解剖结构清晰度打分,评分越高标志着扫描区域可读性好。评分者6周后再次重复随机评分,总的图片质量是所有位点评分之和。采用组内相关系数(ICC)确定1轮和2轮间观察者内和观者间的信度,采用多个观察者加权的kappa系数来分别评价观察者内和观察者间的信度,采用Wilcoxon符号秩检验分析不同参数间的差异。四、研究结果第一部分CHARITé与PRODISC?-L两种假体金属盖板在1.5T磁共振主磁场中的最大偏移角度分别为6.0±0.24°和7.5±0.27°(P<0.01),两者均远小于45°,表明两种假体盖板在1.5T磁共振主磁场中收到的磁力小于自身重力;在磁共振检查前后尸体标本上植入物附近温度升高分别为:CHARITé组为0.8℃(对照处升高0.4℃,P<0.01) , PRODISC?-L0.9℃(对照处升高0.3℃,P<0.01);结果表明植入物处在磁共振检查射频波作用下能够产热,但在尸体标本上产热使邻近组织温度升高均小于1.0℃。第二部分设计的水模系统能够用来评价伪影大小,各序列下最大截面的伪影形态如图2-5所示,最大直径如表2-2所示,两种植入物在各扫描序列矢状位及横切位图像中的伪影均影响邻近结构辨识, TSE、STIR和Turbo Dark Fluid序列相对TSE脂肪饱和、Flash和SE序列伪影较小。磁共振成像带宽、视野、相位编码方向、回波链长度等参数均明显影响伪影大小。第三部分每个观察者第1轮和2轮的观察者内信度为优,平均0.85(范围从0.63-0.97)。总得分的观察者间平均ICC第一轮为0.84,第二轮为0.76,评分系统总的信度为优。对每个位点,观察者内的信度由加权的kappa值决定。数据表明在大多数位点的评分中观察者内的一致性均较好。优化参数前后的MRI图像质量进行比较,矢状位图像优化前得分为:8.45±1.41,优化后得分提高至15.75±2.34(P<0.01),横切位图像优化前得分为:8.63±1.94,优化后得分提高至13.33±1.30(P<0.01),总体上优化参数明显提高了图像植入物区结构的清晰度。四、研究结论1. CHARITé与PRODISC?-L腰人工椎间盘在在1.5T磁共振检查环境中受力和产热两项指标在安全性范围内,腰人工椎间盘置换术后患者可以安全进行磁共振检查。2.建立的水模能够实现对金属伪影的定量评价。常规SE,TSE, Flash, STIR, TSE脂肪饱和等序列扫描发现两种植入物在矢状位及横切位均影响邻近结构辨识, TSE,STIR和Turbo Dark Fluid序列相对TSE脂肪饱和, Flash与SE序列金属伪影较小。带宽、视野、相位编码方向、回波链长度等均影响伪影大小。3.对TSE序列进行参数优化(增加带宽、减少视野、增加回波链长度)能够大大减少金属伪影,增加腰人工椎间盘植入术后磁共振图像的诊断性能。更多还原

【Abstract】 1. BackgroundTotal disc replacement (TDR) is a new alternative for the treatment of lumbar degenerative disc disease. Compared with conventional spinal fusion surgery, disc arthroplasty can maintain the motion at the affected segment. Therefore it is supposed to recover normal lumbar function and reduce postoperative degeneration at adjacent disc levels, which was reported as a common complication after spinal fusion surgery.Despite the great perspective with TDR, a few problems remain to be investigated. Some studies have found that facet joints degeneration following TDR were accelerated. And same to other spinal surgeries, some complications can happen such as insufficient decompression, neurological deficit etc. Another potential problem following TDR surgery is, as patient undergone TDR is usual younger than that of spinal fusion surgery, the morbidity of other spinal disorders is increased during much longer follow-up after surgery. So there is a high likelihood for patients undergone TDR to use MRI to evaluate all these potential complications and spinal disorders.Magnetic resonance imaging (MRI) is a very useful diagnosis tool for spinal disorders as its excellent soft tissue contrast. But when a patient with metallic implant is undergoing MR imaging procedure, two issues must be concerned: one is whether the patient is safe under MR environment because of the attraction force induced by ferromagnetism material under the powerful magnetic field environment of MRI and heating effect of the radio frequency (RF) during MR procedure. The other matter is the metal artifacts on image of MRI,which usually impair diagnostic information. recognized the need for standards to address the safety of implants and other medical devices in the MR environment.On the safety and compatibility of implant materials and medical devices in the magnetic resonance (MR) environment,American Society for Testing and Materials (ASTM International) began developing standard test methods. To date, five ASTM standards addressing testing and marking medical devices and other items for use in the MR environment have been published.Many strategies can be employed to reduce the size and scale of such artifacts; factors include selection of metal hardware material, patient positioning, and MRI sequence adjustments and techniques. In general, gradient recalled echo acquisitions should be avoided in the setting of bulk orthopedic hardware. To reduce deleterious metal artifact related to bulk metallic devices, high spatial resolution fast spin echo sequences with minimal interecho spacing and increased receiver bandwidth should be used. Orientation of the frequency and phase encoding gradients should be selected such that misregistration artifacts arising from the metal are directed away from areas of anticipated clinical diagnostic interest. Additionally, newer imaging sequences, such as view angle tilting, can be adopted in an imaging strategy to minimize effects of metal artifact. Inhomogeneity of fat suppression during MRI in the setting of metal can be minimized by use of inversion recovery sequences, rather than frequency selective fat saturation. Most adjustments made to reduce paramagnetic susceptibility artifacts result in reduction in the image quality overall, usually the result of reduced signal-to-noise ratio. MRI quality in the region of metal can be improved significantly with application of these techniques, reducing metal-related artifacts and enhancing diagnostic assessment of structures in the environment of orthopedic hardware.Although some authors have used MRI for postoperative evaluation of TDR, no any study has been reported on the safety data and metal artifacts distribution of lumbar artificial disc during MR imaging with different sequences. The present study was to determine the safety data and metal artifacts of two artificial disc devices in MR imaging by a few control experiments.2. Objectives(1) To evaluate the safety of two artificial disc devices under 1.5T MR imaging procedure by measuring the attraction force of the metal device in main magnetic field and the quantity of heat of them in body specimen during MR imaging procedure.(2) To analysis quantitatively the size of metal artifacts of two artificial disc devices by a self-design water phantom and to compare the artefact sizes of them under different sequences and parameters in MRI.(3) To study the effect of sequence with optimized parameters on the quality of MR imaging after total disc replacement.3. Materials and methodsPartⅠSafety of CHARITéand PRODISC?-L artificial disc devices during MRI procedureAccording to the recommended test standard of ASTM F2052-06el, an angle-measurement instrument was designed to measure the deflection angle of the endplates of two artificial discs in the static magnetic field of MR system. Suspended with a thin nylon thread, the superior endplates of two disc prosthesises were individually placed at the entrance of magnetic field, where the greatest displacement was produced.According to the recommended surgical procedure by device manufacturer, artificial disc was placed in L5/S1 intervertebral space of a human cadaver. The temperatures on the adjacent tissue of the implant and L4/L5 intervertebral disc (as a control) were measured respectively using a digital probe thermometer. Then MR scan was performed using a body coil and turbo spin echo sequence (TSE) ( scan time was 30 minutes, which represented“a worse condition”for heating effect of RF in clinical condition). When the scan was over, the temperature of the same sites as that prescan was measured again.PartⅡDesign of a water phantom used for evaluation of metal artifacts and analysis the size of metal artifacts of two artificial disc devices under different sequences and parameters in MRI.In order to evaluate the metal artifacts of two artificial discs under different MR sequences , we designed a rectangular water phantom according a method modified from ASTM F2119-01, which was made of polymethyl methacrylate plastics. The water phantom was filled with CuSO4 solution (1 g/L). The artificial disc was then tied and suspended with three thin nylon threads to keep in the centre of the water phantom. MR scan was performed using SE(Spin Echo, T1-weighted) ,TSE(Turbo Spin Echo ,T1/T2 -weighted), Flash(Incoherent Gradient Echo ), STIR(Short T1 Inversion Recovery), TSE fat saturation and Turbo Dark Fluid (Long Tau Inversion Recovery) sequences respectively and different parameters(field of view, direction of phase encoding, excitation time, echo time, flip angle, saturation zone, bandwidth, echo train length, et al ). Then the images were obtained and analysised. OSIRIS software was used to measure the maximal diameter of metal artifacts.PartⅢApplication of the sequence with optimized parameters in MR imaging for patients after total disc replacement.The two artificial discs were placed in L5/S1 intervertebral space of a human cadaver respectively. MR procedure with and without optimized parameters were performed for the cadaver to compare the quality of image. Typical images in the same slice of both saggital and transverse cross-section were saved as a JPEG format documentation, which were numbered and sent to six spinal experts or radiologists. The MRI quality was graded using the Jarvik scale. A higher score corresponded to a better visualization of that particular area on the MR scans. The examiners repeated scoring after a second randomization, 6 weeks later. The overall quality score was the sum of the individual scores for all sites. The intraclass correlation coefficient (ICC) was used to determine intraobserver and interobserver reliability between rounds 1 and 2. For each site score, the weighted kappa coefficient and kappa coefficient for multiple observers were used to assess the intraobserver and interobserver reliabilities, respectively. To look for differences in scoring between sequences with and without optimized parameters, analysis of Wilcxon signed-rank test was used.4. ResultsPartⅠThe maximal deflection angle under MR static field was 7.5±0.27°for the endplate of CHARITéprosthesis and 6.0±0.24°for the endplate of PRODISC?-L prosthesis (P<0.01). Both wereless than 45°, which mean the magnetically induced deflection force is less than the force on the device due to its weight.; The temperature rise on the adjacent tissue of the two artificial disc duo to the implants were 0.8 and 0.9℃for CHARITéand PRODISC?-L prosthesis respectively. The results also showed that even under the worse condition. The results of the MRI-related heating tests indicated that the temperature change measured for the two implants were less than 1℃and could be well tolerated.PartⅡThe water phantom which was designed by us could be used for evaluation of metal artifacts. The images of two artificial disc with T2-weighted TSE sequence were shown in Figure 2-5 and the size of metal artifacts in z-direction for axial image and x- direction for sagittal image under different sequences were shown in Table 2-2. The size of metal artifacts on images of TSE(T1/T2–weighted),STIR and Turbo Dark Fluid sequences were relatively less than those of TSE fat saturation, Flash and SE (T1-weighted) sequences. Different parameters(field of view, direction of phase encoding, echo time, flip angle, saturation zone, bandwidth, echo train length, et al ) had significant effect on the extent and size of metal artifacts.PartⅢThe intraobserver reliability for rounds 1 and 2 for each examiner was excellent, with mean 0.85 ranging from 0.63 to 0.97. The mean interobserver ICC for the total score was 0.84 and 0.76 for rounds 1 and 2, respectively. The total reliability of the scoring system was moderate. The weighted kappa statistic determined the intraobserver reliability for each MRI site since the ordinal scaling system was used in this study. It shows that there was substantial intraobserver agreement in scoring most sites. When comparing the scores of sequences with and without optimized parameters, the saggital image with optimized parameters was scored 8.45±1.41 while the image without optimized parameters was 15.75±2.34(P<0.01) ; For trans-section images the scores were 8.63±1.94 and 13.33±1.30 respectively. In general, parameter optimization enhanced diagnostic assessment of structures in the environment of orthopedic hardware.5.Conclusions(1) According our experiments, because of the minor magnetic field interactions and heating effect of RF, MRI may be performed immediately after these devices are implanted. Based on the results of the tests , the CHARITéand the PRODISC?-L artificial disc will not present an additional hazard or risk to a patient undergoing an MRI procedure using a scanner operating with a static magnetic field of 1.5T or lower and un der the MRI-related heating conditions used for this evaluation.(2) Metal artifacts of the two implants are similar in size and may present problems if the anatomical region of interest is in or near the area where these implants are located (e.g., vertebral canal at affected segment). The size of metal artifacts on images of TSE(T1/T2–weighted),STIR and Turbo Dark Fluid sequences were relatively less than those of TSE fat saturation, Flash and SE (T1-weighted) sequences.(3) The TSE sequence with optimized parameters ( increasing BW and EDL, reducing FOV) could enhance greatly the diagnostic performance of MRI for patients after TDR.更多还原