Patient-Specific Technology for In Vivo Assessment of 3-D Spinal Motion

摘要 目的：本研究的长期目标是设计一种非侵入性的方法来准确评估3D， 体内脊柱运动学，以促进旨在改善诊断和治疗评估的临床研究 治疗脊椎退化 假设：我们的中心假设是，可以开发一种基于MRI的患者特异性技术， 精确地量化连续的3D椎骨运动、小关节运动以及椎管和椎间孔 利用临床医生可以随时获得的工具，并且没有过度的辐射暴露。 上述假设是基于我们先前在实验室的工作，这导致了一种 用于人体尸体重建手术研究的骨特异性三维运动学评估工具。 具体目的：（1）开发和体外验证非侵入性双平面荧光透视成像 脊椎解剖学配准技术。工作假设1：双平面透视配准 可以设计一种方法，结合三维MRI解剖模型，可以准确地评估节段性 在离散位置的运动学与利用不透射线球体的实验室技术一样精确 植入在每个椎骨上(2)用于预测的算法的开发和离体验证 连续运动学数据来自在离散姿势中测量的节段运动。工作假设2： 使用三维样条函数插值可以创建基于连续运动学数据的精确预测 基于针对离散位置获得的双平面荧光透视数据。 研究计划：我们将结合对人类尸体脊柱的实验研究，CT和MRI， 基于解剖模型和双平面透视成像，以开发一种非侵入性方法， 连续的三维运动学数据中指定的目的，拟议的研究。我们的总体方法是 如流程图所示。研究的第一步是：（1）设计/完善一个算法，用于配准 使用双平面荧光透视成像的椎骨解剖结构;以及（2）设计/改进插值算法， 从在离散位置处采集的椎骨运动数据预测连续数据。这些算法将是 在人尸体脊柱标本上收集的实验数据上实施。尸体实验 将同时产生：（i）使用不透射线的方法获得的“黄金标准”连续运动学数据， 植入椎体的球体，以及由红外光组成的光电跟踪系统 连接到每个椎骨和IRED跟踪相机的发光二极管，以及（ii）从椎骨和IRED跟踪相机导出的运动学数据。 提出的双平面配准技术结合插值算法， 数据我们将使用上面收集的实验数据来构建和比较3种特定于生物体的模型：（1）a 基于CT的黄金标准模型，（2）使用双平面荧光透视图像配准的基于CT的模型，以及（3） 使用双平面荧光透视图像配准的基于MRI的模型。运动学参数计算如下： 3D数据将产生关于所提出的非侵入性技术的准确度和精确度的信息。 意义：拟议研究的贡献将包括：（1）一个强大的，非侵入性的运动学 利用临床医生可以随时获得的工具的评估程序;和（2）非侵入性方法 在没有过度辐射的情况下获得连续的3-D运动学数据;这与患者的 三维解剖模型将允许动态评估旋转轴、小关节运动、椎管和 椎间孔狭窄和功能负荷情况下的体内三维椎间盘变形。

Abstract Purpose: The long-term goal of this research is to devise a non-invasive method to accurately assess 3-D in vivo spine kinematics to facilitate clinical studies aimed at improving diagnosis and assessment of treatments for degenerative spine conditions. Hypotheses: Our central hypothesis is that a patient-specific, MRI-based technique can be developed to accurately quantify continuous 3-D vertebral motions, facet joint motions, and spinal canal and foraminal stenosis utilizing tools that could be readily available to clinicians, and without excessive radiation exposure. The above hypothesis is based on our prior work in the laboratory, which has led to the development of a specimen-specific, 3-D kinematic assessment tool for human cadaveric studies of reconstructive surgeries. Specific Aims: (1) Development and ex vivo validation of a non-invasive, bi-planar fluoroscopic imaging technique for registration of vertebral anatomy. Working hypothesis 1: A bi-planar fluoroscopic registration method can be devised that, in conjunction with a 3-D MRI anatomic model, can accurately assess segmental kinematics at discrete positions as accurately as the laboratory technique that utilizes radiopaque spheres implanted on each vertebra. (2) Development and ex vivo validation of an algorithm for prediction of continuous kinematic data from segmental motions measured in discrete postures. Working hypothesis 2: Interpolation using 3-D spline functions can create an accurate prediction of continuous kinematic data based upon the bi-planar fluoroscopic data obtained for discrete positions. Research Plan: We will use a combination of experimental studies on human cadaveric spines, CT- and MRI- based anatomic models and bi-planar fluoroscopic imaging to develop a non-invasive method to acquire continuous 3-D kinematic data as specified in the aims of the proposed study. Our overall approach is illustrated in a flowchart. The first steps of the study are to: (1) design/refine an algorithm for registration of vertebral anatomy using bi-planar fluoroscopic imaging; and (2) design/refine an interpolation algorithm to predict continuous data from vertebral motion data acquired at discrete positions. These algorithms will be implemented on experimental data collected on human cadaveric spine specimens. The cadaveric experiments will simultaneously yield: (i) "Gold Standard" continuous kinematic data that is obtained using radiopaque spheres implanted on vertebral bodies, and an optoelectronic tracking system consisting of infrared light emitting diodes attached to each vertebra and IRED tracking cameras, and (ii) kinematic data derived from the proposed bi-planar registration technique in conjunction with an interpolation algorithm to derive continuous data. We will use the experimental data collected above to build and compare 3 specimen-specific models: (1) a CT-based gold standard model, (2) a CT-based model using bi-planar fluoroscopic image registration, and (3) an MRI-based model using bi-planar fluoroscopic image registration. Kinematic parameters calculated from the 3-D data will yield information on the accuracy and precision of the proposed non-invasive techniques. Significance: The contributions of the proposed study will include: (1) a robust, non-invasive kinematic assessment procedure utilizing tools that could be readily available to clinicians; and (2) a non-invasive method to acquire continuous 3-D kinematic data without excessive radiation; which in conjunction with the patient's 3-D anatomic model will allow dynamic assessment of axes of rotation, facet joint motions, spinal canal and foraminal stenosis, and in vivo 3-D disc deformations under functional loading scenarios.