Computer Assisted Ultrathin Brochoscopy

描述（由申请人提供）：这个拟议的成像系统应用开发的工业-学术合作伙伴关系的目的是开发用于计算机辅助超薄支气管镜（CAUB）活检小周围肺结节的新方法。这些传统上很难用标准的柔性支气管镜诊断。我们得出结论，最近的技术发展，包括：超薄支气管镜的引入；能够产生全肺近各向同性成像的多探测器CT扫描仪的广泛使用；新开发的先进图像处理算法，包括气道和血管分割和骨架化的复杂方法，以及虚拟内窥镜，结合起来有可能改变目前诊断肺结节的临床算法。虽然迄今为止人们的注意力几乎完全集中在使用虚拟支气管镜（VB）来评估中央气道，但我们认为需要一种更可靠的方法来识别周围气道，以扩大图像引导支气管镜的范围。该应用的核心是开发自动化方法，用于获取通往小周围肺结节的虚拟内镜路径，包括在使用高分辨率近各向同性CT图像无法识别病变附近支气管的情况下，使用肺动脉作为周围支气管位置的替代标记。这些自动化技术可以与超薄支气管镜相结合，从而可以常规诊断以前认为常规柔性支气管镜无法诊断的肺结节。具体目标1将是创建从中央气道到计算机生成的小周围肺结节的自动虚拟内镜路线图，作为评估自动路径规划和气道/动脉对应的可行性的一种手段。具体目标2将在试点项目中讨论这些技术的潜在临床应用。预计这些结果将使超薄支气管镜检查成为评估周围肺结节的首选程序，从而改变目前的临床范例。

DESCRIPTION (provided by applicant): The purpose of this proposed Industry-Academic Partnerships for Development of Imaging Systems Application is to develop novel approaches for computer assisted ultrathin bronchoscopy (CAUB) for biopsying small peripheral lung nodules. These traditionally have been difficult to diagnose using standard flexible bronchoscopes. We conclude that recent technical developments, including: the introduction of ultrathin bronchoscopes; widespread availability of multidetector CT scanners capable of generating near isotropic imaging of the entire lung; and newly developed advanced image processing algorithms, including sophisticated methods for segmenting and skeletonizing airways and vessels as well as virtual endoscopy, combined have the potential to alter current clinical algorithms for diagnosing lung nodules. While to date attention has almost exclusively focused on the use of virtual bronchoscopy (VB) to evaluate only the central airways, in our opinion a more robust approach to identifying peripheral airways is needed to extend the range of image guided bronchoscopy. At the heart of this application is the development of automated-methods for obtaining virtual endoscopic pathways leading to small peripheral lung nodules, including the use of pulmonary arteries as surrogate markers for the location of peripheral bronchi in those cases in which bronchi cannot be identified adjacent to lesions even using high-resolution near-isotropic CT images. These automated techniques can then be coupled with ultrathin bronchoscopy allowing the routine diagnosis of pulmonary nodules previously considered inaccessible to routine flexible bronchoscopy. Specific Aim 1 will be to create automatic virtual endoscopic roadmaps from central airways to small computer generated peripheral lung nodules as a means to assess the feasibility of automated pathway planning and airway/arterial correspondence. Specific Aim 2 will address the potential clinical application of these techniques in a Pilot Project. It is anticipated that these results will allow ultrathin bronchoscopy to emerge as the initial procedure of choice for assessing peripheral lung nodules thus altering present clinical paradigms.