Imaging, Guidance, and QA for Emerging High-Precision Neurosurgical Techniques

项目摘要/摘要 新兴的神经外科技术在治疗越来越多的运动频谱方面提供了潜在的突破 精神障碍和痴呆症(包括阿尔茨海默氏症、抽动症、自闭症、抑郁症，甚至肥胖)。这些 新出现的手术方法通过以下方式延长了深部脑刺激(DBS)在帕金森病中的既定成功 使用新的电极刺激器经脑室传递到下丘脑周围的靶点。在内窥镜检查时 入路提供了可靠的进入脑室的途径，这种途径会导致脑脊液(CSF)的损失和脑移位 对于这些新的DBS疗法，在非常感兴趣的区域最大可达~10毫米。因此，实现这样的好处 前景看好的技术需要神经导航领域超越最先进水平的进步。此外，小说的使用， 这种技术中的定向电极需要一种方法来指导和确认刺激器的放置。尤其是在 在这类新疗法开发的早期阶段，解决与几何精度有关的不确定性是很重要的 以区别于潜在的神经生理学和其他可能影响安全性和预后的因素。 我们建议发展和评估术中成像、配准和指导方面的以下进展 实现机器人辅助脑室镜进路的平台，以克服 神经导航的传统局限性，并支持新兴的新一代DBS疗法： (目标1)开发高质量的术中锥束CT(CBCT)，使用三维图像重建方法，推动 图像质量超过CBCT的常规限制，提供足以驱动可变形配准的图像质量 术前核磁共振，精确定位刺激器位置，并提供对并发症/颅内出血的检查。 (目的2)开发3D-2D图像配准方法以将低剂量术中X线片与：(A)术前MRI用于 自动患者登记；和(B)DBS电极(包括新型定向刺激器)的参数模型，用于 指导和确认刺激器的放置，其精确度和精确度超过常规跟踪。 (目的3)发展多模式可变形图像配准(MR-CBCT)以解决术前MRI和 术中CBCT-特别是脑脊液流出后脑室周围深部变形-使用一种快速的方式- 用于将MRI/规划数据准确转换为CBCT和内窥镜的不敏感、微分同胚层析的Demons方法。 (目标4)开发内窥镜视频配准，以便在内窥镜中直接呈现3D图像和规划信息 场景，在脑室镜手术过程中提供目标和关键结构的准确可视化。 (目标5)将目标1-4中的方法转化为临床研究，以便在现实条件下定量评估绩效 条件，并结合在一个集成系统中的机器人辅助脑室镜(RAV)方法，以达到DBS的目标。 该提案推进了之前在颅底和脊柱神经外科的工作，提供了成功的高可能性，使 下一代DBS，并创建了一种超越最先进水平的图像引导手术机器人集成系统- 在未来的学术和行业合作伙伴关系下，为临床系统的开发和翻译提供了有价值的试验台。

PROJECT SUMMARY / ABSTRACT Emerging neurosurgical techniques offer potential breakthroughs in treatment of a growing spectrum of movement disorders and dementia (including Alzheimer’s disease, Tourette’s syndrome, autism, depression, and even obesity). These emerging surgical approaches extend the established success of deep-brain stimulation (DBS) in Parkinson’s disease by using novel electrode stimulators delivered trans-ventricularly to targets about the hypothalamus. While endoscopic approach provides reliable access to the ventricles, such access imparts a loss of cerebrospinal fluid (CSF) and brain shift up to ~10 mm in the very regions of interest for these novel DBS therapies. Therefore, realizing the benefit of such promising techniques requires advances beyond the state of the art in neuro-navigation. Moreover, the use of novel, directional electrodes in such techniques requires a means to guide and confirm stimulator placement. Especially in the early stages of development of such novel therapies, it is important to resolve uncertainties related to geometric precision in order to differentiate from underlying neurophysiology and other factors that may affect safety and outcome. We propose to develop and evaluate the following advances in intraoperative imaging, registration, and guidance to realize a platform for robot-assisted ventriculoscopic approach to deep-brain targets in a manner that overcomes conventional limitations of neuro-navigation and supports the emerging generation of novel DBS therapies: (Aim 1) Develop high-quality intraoperative cone-beam CT (CBCT) using 3D image reconstruction methods that propel image quality beyond conventional limits of CBCT, providing image quality sufficient to drive deformable registration with preop MRI, precisely localize stimulator placement, and provide a check against complication / intracranial hemorrhage. (Aim 2) Develop 3D-2D image registration methods to relate low-dose intraoperative radiographs with: (a) preop MRI for automatic patient registration; and (b) parametric models of DBS electrodes (including novel directional stimulators) for guidance and confirmation of stimulator placement with precision and accuracy beyond that of conventional tracking. (Aim 3) Develop multi-modality deformable image registration (MR-CBCT) to resolve alignment between preop MRI and intraoperative CBCT – particularly peri-ventricular deep-brain deformation following CSF egress – using a fast, modality- insensitive, diffeomorphic Demons method for accurate transformation of MRI / planning data to CBCT and endoscopy. (Aim 4) Develop endoscopic video registration to render 3D image and planning information directly in the endoscopic scene, providing accurate visualization of target and critical structures during ventriculoscopic approach. (Aim 5) Translate the methods from Aims 1-4 to clinical studies for quantitative evaluation of performance under realistic conditions, and combine within an integrated system for robot-assisted ventriculoscopy (RAV) approach to DBS targets. The proposal advances previous work in skull base and spinal neurosurgery, offering a high likelihood of success in enabling next-generation DBS, and creates an integrated system for image-guided surgical robotics beyond the state of the art – a valuable testbed for development and translation of clinical systems under future academic-industry partnership.