【作者】 吴冰；

【导师】 李明；

【作者基本信息】 第二军医大学 ， 外科学， 2012， 博士

【摘要】 背景：脊柱侧凸是一种复杂的三维畸形，以特发性脊柱侧凸最为常见，约占全部脊柱侧凸的80%，好发于青少年，文献报道其发病率高达0.75%～2.4%。对于侧凸角度较小的患者，通常可采用支具进行保守治疗，而对于严重脊柱侧凸的患者手术矫形是唯一有效的治疗方法，但脊柱侧凸矫形术是脊柱外科手术的难点之一。后路全椎弓根螺钉内固定矫形术成为脊柱侧凸矫形手术的最常用术式，然而脊柱侧凸的椎弓根螺钉植入具有极高的难度和风险。文献报道胸椎椎弓根螺钉的误置率高达6%-54.7%。目前常用的椎弓根螺钉安全植入辅助方法包括手术技术的革新，如“漏斗技术”；手术器械的改良；常用的是术中X线透视，但该方法仅能显示椎弓根的二维平面图形，影像质量较差，容易出现失误，术中医患双方均受到射线的辐射。计算机辅助导航技术可显著降低误置并发症，但也有配准误差，产生不正确的导航信息，导致钉道建立失败；此系统设备昂贵，操作复杂，影响其临床应用与推广。神经电生理监测可以有效预防神经系统并发症的发生，但无菌要求高，需要手术医生和麻醉医生的术中配合，需要专业技术人员的操作和分析判断，临床应用未普及。超声具有无电离辐射、费用低廉、安装携带方便的优点。Mujagic等在水槽中，使用1MHz和3.5MHz两种频率的超声探头测得超声信号在人椎体松质骨内的穿透深度为1.65-1.9cm。Raphael等使用频率为2.5MHz超声在水槽中对羊椎体松质骨成像研究，发现在椎弓根径路内距离皮质骨1.5cm处可观察到振幅增高的波形。国内文献未见此类研究的报告。上述作者的研究可满足超声导航的基本要求。本实验为脊柱椎弓根置钉超声导航做数据准备和基础实验。本课题受国家自然科学基金资助。项目批准号：81101396。目的：1.应用超声透射法明确脊柱松质骨的超声参数。2.应用超声背散射法明确超声信号在椎体松质骨内的穿透深度。3.应用超声时域有限差分仿真对照，探讨超声信号对椎体松质骨-皮质骨界面预判的波形特点和预判距离4.探讨对椎体松质骨-皮质骨界面预判具备优势中心频率的探头。方法：所有实验在内壁衬有吸声材料的恒温水槽中进行，水温控制在20℃左右，水槽中加入蒸馏水及适量表面活性剂，激发探头发射超声信号，接收信号经超声分析仪放大，显示于示波器，信号输入PC机，应用MATLAB软件整理分析数据，重建波形。1.超声透射实验：超声信号分别由0.5MHz、1MHz、1.5MHz、2.25MHz、3.5MHz中心频率的激发探头发射，透射通过七枚腰椎松质骨样本，厚度为0.9cm，由接收探头接收信号，分析图形特点，测量中心频率衰减、宽带频率衰减和声速。2.椎体穿透深度实验：超声信号分别由0.5MHz、1MHz、1.5MHz、2.25MHz、3.5MHz中心频率的探头发射，透射通过不同厚度的人体七枚腰椎松质骨样本，经由钢板反射，再次透射通过骨样本，由同一探头接收信号。分析图形特点，测量样本的背散射穿透深度。3.脊柱超声导航基础研究实验：超声信号分别由1MHz、1.5MHz、2.25MHz、中心频率的探头发射，透射通过不同厚度的人体七枚腰椎椎体骨样本，经由皮质骨反射，再次透射通过骨样本，由同一探头接收信号。分析图形特点，获取有效波形及距离皮质骨的距离。超声时域有限差分仿真实验：一枚腰椎椎体骨样本行Micro-CT扫描，截取二维图像，建立时域有限差分仿真模型，设置初始条件、边界条件和激励条件，完成有限差分仿真，获取仿真超声背散射信号。结果：1．超声透射实验：七枚人第三腰椎椎体松质骨样本，应用0.5MHz、1MHz、1.5MHz、2.25MHz、3.5MHz超声探头，每个样本选取五个感兴趣区，应用超声透射法测量声学数据。中心频率衰减为7.68±4.06dB/cm (0.5MHz)，18.86±7.49dB/cm (1MHz)，29.96、±11.23dB/cm (1.5MHz)，41.40±21.07dB/cm(2.25MHz)，52.96±15.33dB/cm (3.5MHz)；归一化宽带超声衰减（nBUA）为28.70±6.37dB/(cm MHz)(0.5MHz)，17.31±10.39dB/(cm MHz)(1MHz)，20.70±6.94dB/(cm MHz)(1.5MHz)，16.52±5.35dB/(cm MHz)(2.25MHz)，12.46±5.16dB/(cmMHz)(3.5MHz)；声速为1514.41±16.81m/s(1501.67-1506.30m/s)，(0.5MHz)；1508.61±4.97m/s，（1502.03-1518.23m/s），(1MHz)；1502.56±2.19m/s，（1500.20-1507.02m/s），(1.5MHz)；1500.60±0.59m/s，（1500.04-1502.44m/s），(2.25MHz)；1500.15±0.15m/s，（1500.00-1500.61m/s），(3.5MHz)。统计分析：组间比较中心频率衰减在各组间均有显著性差异；nBUA：显示1、1.5、2.25MHz三组间无显著性差异，在其余各组间均有显著性差异；SOS：各组间均有显著性差异2．以钢板为反射的超声背散射穿透距离实验：七枚人第四腰椎椎体松质骨样本，应用0.5MHz、1MHz、1.5MHz、2.25MHz、3.5MHz超声探头测量超声信号在松质骨内的穿透距离2.19±0.34cm (0.5MHz)，1.17±0.38cm (1MHz)，0.93±0.44cm (1.5MHz)，0.69±0.20cm (2.25MHz)，0.69±0.20cm (3.5MHz)。统计分析各组间比较无显著性差异。3、超声仿真与椎体背散射实验：超声FDTD仿真，在距离腹侧皮质骨0.2cm处出现振幅数倍增高于邻近波形的仿真超声信号，振幅升高前的距离内振幅波动较大。实验研究七枚腰五椎体松质骨样本，距离皮质骨0.6cm距离处的时域波形图振幅，整体趋势显著升高，在1MHz和1.5MHz振幅有一倍以上增长，2.25MHz未达到一倍增长。振幅升高前的距离振幅波动较大。与超声仿真趋势表现一致。将三种不同频率距离1.2cm、0.9cm和0.6cm处的归一化振幅均值统计分析，0.6cm距离处的归一化振幅显著高于1.2cm和0.9cm两组，有显著性差异；1.2cm和0.9cm两组间归一化振幅无显著性差异结论：1.取得了不同频率椎体松质骨的超声参数。超声衰减与频率的关系在1MHz、1.5MHz、2.25MHz，近似呈线性关系。超声衰减与BMD密切相关。声速在椎体松质骨内呈负色散现象。超声透射实验显示1MHz、1.5MHz、2.25MHz的超声探头在椎弓根置入超声导航的探头频率的选择上具有优越性。2.取得了超声背散射法在椎体松质骨的穿透深度数据，以钢板为反射穿透距离有可能高于实际穿透距离。穿透深度与超声频率、BMD、骨微结构密切相关。超声背散射实验显示1MHz、1.5MHz、2.25MHz的超声探头在椎弓根置入超声导航的探头频率的选择上具有优越性。3.超声FDTD仿真可以有效模拟超声信号在松质骨至皮质骨的传导趋势，但二维模型丢失了部分实际三维机构的信息，导致具体数据的偏差。4.在1MHz、1.5MHz两种中心频率探头下，可在距离皮质骨0.6cm距离处出现与邻近信号最低信号振幅有一倍以上增长。1MHz、1.5MHz中心频率探头在松质骨穿透距离和松质骨-皮质骨界面的预判更具优势，适宜应用于超声导航。更多还原

【Abstract】 Background:Scoliosis is a complex three-dimensional deformity. Idiopathic scoliosis is the mostcommon in Scoliosis, accounting for80%of all scoliosis in adolescents. The reportedincidence is up to0.75%to2.4%. Smaller in patients with scoliosis angle, can usuallybe bracing for conservative treatment. The only effective treatment for patients withsevere scoliosis is surgery. Scoliosis surgery is one of the difficulties of spine surgery.Posterior whole pedicle screw internal fixation is the most common procedure inscoliosis surgery. However,the fixation has a very high difficulty and risk. Reported inthe literature, thoracic pedicle screw misplacement rate is up to6%-54.7%.The auxiliary methods for pedicle screw safe implantation include the innovationof surgical techniques, such as the "funnel technique" and the improvement of surgicalinstruments et al. The most common used in clinic is intraoperative X-ray fluoroscopy,but the method can only show the pedicle two dimensional plane, the image of poorquality, prone to errors. Radiation exists between both doctors and patients in peration.Computer-aided navigation technology can significantly reduce the misplacement, butthere are registration errors. Incorrect navigation information leads to the failure of thecorrect screw channel. This equipment is expensive and complicated to operation. It isdifficult to be popularized. Neurophysiological monitoring can be effective in preven-ting the occurrence of neurological complications, but need the high sterile demandand assort between surgeon and anesthesiologist. With the needs of professional opera-ting and analytical judgment, clinical application is not universal.Ultrasonics has advantages of no ionizing radiation, low-cost, easy to install.Mujagic measured the penetration depth of the human vertebral cancellous bone with1MHz and3.5MHz ultrasonic transducers in the sink. The depth is1.65-1.9mm.Raphael observed the ultrasound images of sheep vertebral cancellous bone with2.5MHz transducer in the pedicle path. The increased waveform of amplitude and thedistance of1.5cm between marrow-cortex were observed. Domestic literature has noreport of such research. The study of Mujagic’s and Raphael’s meet the basicrequirement of the ultrasound navigation.The study discusses the feasibility of pedicle screw implantation ultrasoundnavigation and data preparation. The subject is subsidizedby the National Natural Science FundObjective:1. Clear ultrasound parameters of human vertebral cancellous bone with ultrasonictransmitted method.2. Clear the penetration depth of the human vertebral cancellous bone of ultrasoundsignal with ultrasonic backscatter method3. Investigate the ultrasonic waveform characteristics and distance for anticipatingmarrow-cortex interface refering to ultrasonic finite difference time domain(FDTD)simulation.4. Explore the advantaged ultrasound transducer for predicting marrow-cortexinterface.Methods:All experiments were performed in sink of connstant temperarure lining the innerwall of sound-absorbing material being filled with distilled water and the amount ofsurfactant. The water’s temperature was controlled at about20℃. Stimulattingtransducer transmits ultrasound signals. Received signals were amplificated byultrasound analyzer, displayed on the oscilloscope. Signals inputted PC, usedMATLAB software analyzing data.The ultrasound waveform was reconstructed.1. Ultrasound transmitted experiment: the ultrasonic signal were launched from0.5MHz,1MHz,1.5MHz,2.25MHz,3.5MHz center frequency transducer, transmittedthrough seven lumbar cancellous bone samples with the thickness of0.9cm. The trans-mitted signal was received by the receiving transducer. Graphical features were analy-zed. Center frequency attenuation, broadband frequency attenuation and speed ofsound of different frequencies were measured.2. Vertebral penetration depth experiment: the ultrasonic signal were launched from0.5MHz,1MHz,1.5MHz,2.25MHz,3.5MHz center frequency transducer, transmittedthrough seven lumbar cancellous bone samples with different thickness, reflectedthrough the steel plate, transmitted through bone samples again, received by the sametransducer. Graphical features were analyzed. Penetration depth of ultrasound backs-catter was measured.3.Spinal ultrasound navigation basic research experiment: the ultrasonic signal werelaunched from1MHz,1.5MHz,2.25MHz center frequency transducer, transmittedthrough seven lumbar vertebral bone samples with different thickness, reflected throu- gh the vertebral cortex, transmitted through cancellous bone again, received by thesame transducer. Graphical features were analyzed. Penetration depth of ultrasoundbackscatter was measured. Effective ultrasound waveform and the distance to vertebralcortex were obtained.Ultrasonic finite difference time domain(FDTD) simulation: A lumbar vertebralbone sample were scanned by the Micro-CT. The two-dimensional image was interce-pted for the establishment of the FDTD simulation model. The initial conditions,boundary conditions and excitation conditions were setted to complete the FDTDsimulation. The ultrasonic backscatter simulation signal was getted.Results:1. Ultrasound transmitted experiment: Seven of the L3vertebral cancellous bonesamples were transmitted by0.5MHz,1MHz,1.5MHz,2.25MHz,3.5MHz centerfrequency transducer. Each sample were selected five regions of interests(ROIs).Acoustic data were measured. Center frequency attenuation was7.68±4.06dB/cm(0.5MHz),18.86±7.49dB/cm(1MHz),29.96±11.23dB/cm(1.5MHz),41.40±21.07dB/cm(2.25MHz),52.96±15.33dB/cm(3.5MHz);Normalized broadband ultrasound attenua-tion was28.70±6.37dB/(cm MHz,)(0.5MHz),17.31±10.39dB/(cm MHz,)(1MHz),20.70±6.94dB/(cm MHz)(1.5MHz),16.52±5.35dB/(cm MHz)(2.25MHz),12.46±5.16dB/(cm MHz)(3.5MHz); Speed of sound was1514.41±16.81m/s(1501.67-1506.30m/s),(0.5MHz);1508.61±4.97m/s,(1502.03-1518.23m/s),(1MHz);1502.56±2.19m/s,(1500.20-1507.02m/s),(1.5MHz);1500.60±0.59m/s,(1500.04-1502.44m/s),(2.25MHz);1500.15±0.15m/s,(1500.00-1500.61m/s),(3.5MHz). Statistical analysis:center frequency attenuation between the two groups was significant differencesamong all groups. nBUA: For the three groups (1,1.5,2.25MHz),there was no signifi-cant difference, was significant differences between the other groups. SOS: Therewere significant differences between groups2. Ultrasonic backscatter penetration distance experiment with reflection of Steelplate: Seven L4vertebral cancellous bone samples were backscatterred by0.5MHz,1MHz,1.5MHz,2.25MHz,3.5MHz center frequency transducer. The ultrasonic signalpenetration distance in the cancellous bone was2.19±0.0.34cm(0.5MHz),1.17±0.38cm(1MHz)0.93±0.44cm(1.5MHz),0.69±0.20cm (2.25MHz),0.69±0.20cm (3.5MHz). The statistical analysis has no significant difference among all groups.3. Ultrasound simulation and vertebral ultrasound backscatter experiment: In FDTDsimulation, two ultrasonic simulation signals with several times high amplitude thanneighboring waveform amplitude were emerged, which distance to the vertebralcortex is0.2cm. The waveform amplitudes before the distance of0.2cm fluctuated.Seven L5vertebra samples were studied in ultrasound backscatter experiment.Ultrasonic signal located in0.6cm to the diatance to vertebral cortex showed theoverall increased trend significantly, performed high amplitude, which one timehigher than the lowest neighboring waveform amplitude in1MHz and1.5MHzwaveform, but less than one time in2.25MHz waveform. The waveform amplitudesbefore the distance of0.6cm fluctuated. Experimental study was consistent withFDTD simulation. Of three types of different frequency, normalized amplitude dataof the distance in1.2cm,0.9cm and0.6cm to vertebral cortex were analysed.Normalized amplitudes in the distance of1.2cm and0.9cm showed a significantdifference with which in the distance of0.6cm separately. The two groups’ norma-lized amplitude in the distance of1.2cm and0.9cm to vertebral cortex showed nosignificant difference.Conclusions:1. Ultrasonic parameters were acquired in vertebral cancellous bone with differentfrequencies. Ultrasonic attenuation and frequency in1MHz,1.5MHz,2.25MHz,approximately show linear relationship. The ultrasonic attenuation is closely relatedwith BMD. SOS in the vertebral cancellous bone was negative dispersion phenomena.Ultrasound transmitted experiments showed the superiority of the choice of the1MHz,1.5MHz,2.25MHz transducers in the pedicle ultrasound navigation.2. Ultrasonic backscatter penetration depth data in vertebral cancellous bone wasobtained. The depth data with steel plate reflection may be higher than the actualpenetration distance. Penetration depth with the ultrasonic frequency is closely relatedto BMD and bone microarchitecture. Ultrasound backscatter experiments showed thesuperiority of the choice of the1MHz,1.5MHz,2.25MHz transducers in the pedicleultrasound navigation.3. Ultrasound FDTD simulation can effectively simulate the ultrasonic signal in thetrend of cancellous bone to cortical bone conduction. But two-dimensional modelmissed a part of the actual three-dimensional information, leading to the deviation of the specific data.4. Normalized amplitudes in1MHz and1.5MHz center frequency transducers showmore than one time of adjacent minimum signals’ amplitude in which of0.6cmdistance to vertebral cortex.1MHz,1.5MHz center frequency transducers in thecancellous bone penetration depth and anticipated distance to marrow-cortex interfaceshow advan-tage, apply to pedicle screw implantation ultrasound navigation.更多还原