A Novel Imaging Informatics Platform for Craniomaxillofacial Surgery

项目描述（由申请人提供）：本项目的最终目标是开发一个开源的新型成像信息学平台AnatomicAligner，以改进颅颌面（CMF）手术的手术计划方法，从而改善CMF畸形患者的治疗效果。CMF手术包括对颅骨和面部的先天性和后天性畸形的矫正。由于CMF骨骼的复杂性，它需要广泛的术前计划。不幸的是，在过去的50年里，传统的规划方法，如预测跟踪和模拟手术，在石头模型上保持不变。许多不希望的手术结果是这些不完善的方法的结果。为了解决这些问题，我们开发了计算机辅助手术模拟（CASS）系统。虽然还需要很大的改进，但CASS的使用已经消除了传统方法的大部分局限性。不幸的是，它也产生了一个新的问题，即牙合的数字化建立变得更加困难。在矫正先前存在的错咬合或手术重建新咬合的计划过程中，牙齿咬合是一个重要的步骤。目前的黄金标准是使用石牙模型，并在咬合器上手工咬合。不幸的是，在虚拟世界中并非如此。这些牙弓是3D图像。当数字牙齿彼此移动时，它们不会因为碰撞而停止，而是继续相互移动，这在现实世界中是不会发生的。为了彻底解决这些问题，开发一种新的系统是至关重要的，它将集成全自动化的牙齿咬合过程，并显著提高我们的CASS技术。我们的假设是，遮挡可以在计算机规划系统中以数字方式自动建立，并且计算机生成的遮挡与手工拼接一组石头模型（目前的黄金标准）建立的遮挡一样精确。为了证明我们的假设，我们提出了三个具体目标，以开发和验证用于CMF手术的新型成像信息学平台AnatomicAligner。该系统的创新之处在于，医生将首次能够在计算机上高效、准确地计划整个手术，包括自动建立牙齿咬合。新的技术贡献包括：1)一个强大的基于3D分割的方法来实现初始的数字牙齿模型对齐；2)自动化最终数字发音的新方法。该项目的意义在于，AnatomicAligner系统将在CMF规划中产生范式转变。外科医生将能够完全放弃有问题的传统方法，采用更准确、更快、更经济的方法。AnatomicAligner的成功将为CMF手术带来一个新的影像信息学平台。这个平台也可以转化为骨科和其他医学专业。一旦完成，软件（包括源代码和可执行文件）将由研究社区从互联网免费下载。

DESCRIPTION (provided by applicant): The ultimate goal of this project is to develop an open source novel imaging informatics platform, the AnatomicAligner, to improve the surgical planning method for craniomaxillofacial (CMF) surgery and subsequently to improve the treatment outcome of the patients with CMF deformities. CMF surgery involves the correction of congenital and acquired deformities of the skull and face. Due to the complex nature of the CMF skeleton, it requires extensive presurgical planning. Unfortunately, the traditional planning methods, e.g. prediction tracings and simulating surgery on stone models have remained unchanged over the last 50 years. Many unwanted surgical outcomes are the result of these deficient methods. To solve these problems, we have developed a Computer-Aided Surgical Simulation (CASS) system. Although it still needs significant improvements, the use of CASS has eliminated most of the limitations of the traditional methods. Unfortunately, it also creates a new problem that the digital establishment of dental occlusion becomes significantly more difficult. The dental articulation is an important step during the planning process to correct preexisting malocclusions or to surgically reestablish a new occlusion. The current gold standard is to utilize stone dental models and hand-articulate them on an articulator. Unfortunately, the same is not true in virtual world. These dental arches are 3D images. When the digital teeth are moved towards each other, they are not stopped by collision and continue to move through each other, which do not occur in real world. In order to completely solve these problems, it is critical to develop a new system that will integrate fully automated process of dental articulation and significantly improved our CASS technologies. Our hypotheses are that the occlusion can be digitally and automatically established in a computer planning system, and the computer-generated occlusion is as precise as the occlusion established by hand-articulating a set of stone models (the current gold standard). In order to prove our hypotheses, we are proposing three Specific Aims to develop and validate a novel imaging informatics platform, the AnatomicAligner, for CMF surgery. The system is innovative because for the first time, doctors will be able to efficiently and accurately plan the entire surgery in the computer, including automated establishment of dental occlusion. The new technical contributions include: 1) a robust 3D segmentation-based approach to achieve the initial digital dental model alignment; and 2) novel approaches for automated final digital articulation. The significance of this project is that the AnatomicAligner system will produce a paradigm shift in CMF planning. Surgeons will be able to completely abandon the problematic traditional methods for a more accurate, faster and cost effective method. The success of AnatomicAligner will lead to a new class of imaging informatics platform for CMF surgery. This platform can also be transformed to orthopedic surgery and other medical specialties. Once completed, the software (both source codes and executables) will be freely downloaded from internet by research community.