基于三维超声成像和图像配准融合的颅底及颌面外科软组织手术导航关键技术和算法研究

The skull base and maxillofacial soft tissue surgery, especially the treatment of skull base tumors, must be performed accurately and safely. Combination with the digital technology and intraoperative navigation technology is the mainstream and developmental trend in this international research field. Through the early 10 years of research, our group researched and developed independently the highly precise optical navigation hardware and software system for craniomaxillofacial skeletal deformities, which was in industrialization after it got the SFDA and CE certification, and has been widely used in the bone diseases. But due to the characteristics of soft tissue deformation that results in image drift during the operation, while the lack of real-time measurement method of the deformation, it is easy to produce large navigation error, which has become the development bottleneck of this leading technology. Therefore, this research project will be carried out in 1) the algorithm and key technology in registration and fusion of the ultrasonic image data and CT/MR data. 2) the establishment of real-time communication among the ultrasound equipment, image registration system and the optical navigation system, in order to monitor the soft tissue deformation, and improve the positioning accuracy of soft tissue. 3) the independent research and development in surgical navigation hardware and software system for skull base and maxillofacial soft tissue operation, based on three dimensional ultrasound imaging technology, and the complete the experimental research of soft tissue operation. Aims at the clinical practical problems and the needs of industrialization, the success of our research will be a great breakthrough and innovation in the field of digital craniomaxillofacial surgery.

颅底及颌面软组织外科，特别是颅底肿瘤的治疗，必须施以精准、安全的定位手术操作，与数字化及术中导航技术相结合是本领域国际主流和前沿发展趋势。本课题组通过10余年的前期研究，自主研发了高精度光学颅颌面骨畸形整复导航硬软件系统，经产业化（SFDA、CE认证）已在国内外临床对骨组织疾病广泛开展应用。但由于软组织易发生形变的特性从而造成术中影像漂移，同时又缺乏实时测量其形变的手段，易产生较大的导航定位误差，成为该领先技术的发展瓶颈。由此，本项目将开展1）超声影像数据与CT/MR等多模数据配准、融合、驱动的算法和关键技术研究；2）建立超声设备、配准图像与光学导航系统的实时通讯；监测术中软组织形变，提高对软组织的定位精度。3）自主研发基于三维超声成像驱动的颅底及颌面软组织手术导航硬软件平台并完成实验手术研究。旨在着重针对临床重大实际问题和产业化需求，其实现将是数字化颅颌面外科领域的重大突破和创新。

颅颌面部由于解剖结构复杂性，手术难度及风险高，特别是对于颅底及软组织肿瘤的治疗，给外科医生带来很大的挑战，这要求外科手术必须以最小创伤完成精确的术中定位操作。目前在临床上广泛使用的高精度硬组织手术导航系统在手术执行精度和精准定位等方面已经无法满足治疗需求及专业发展，并且增加了并发症的产生。.针对上述难题，本项目完成了以下工作:.1、建立了超声与CT图像标定、采集及图像预处理的算法。.2、自主建立了超声与CT/MR等多模数据的配准融合算法 - 基于局部梯度模式（mi-LDP）算法，通过测量图像融合精度共获得13 组误差数据，最小误差为1.6mm，最大误差为3.08mm，平均误差为2.30±1.1mm；记录其图像融合所需平均时间20.9s,明显小于目前主流配准算法所需时间35.2s。.3、自主开发了基于超声等多模数据的颅颌面外科智能机器人规划及控制软件，实现了超声设备与光学导航系统的实时通讯，通过光学配准将术中患者实时的坐标信息与医学图像信息进行匹配，用于软组织导航手术。该导航软件包括配准、监测、硬件控制、超声图像处理及硬件接口等功能模块。 .4、制作建立了基于三维超声成像和图像配准融合技术的颅底及颌面外科软组织手术导航系统原理样机。完成了系统的结构设计图以及数据采集和处理的流程图，并对导航硬软件系统进行性能测试。在超声成像的辅助下利用导航系统对实验对象进行手术，评价了术中超声图像数据用于导航定位的可操作性和精确性。 . 本项目提交的基于三维超声成像和图像配准融合技术的颅底及颌面外科软组织手术导航系统原理样机，经过硬软件系统的兼容性测试及精度测试，初步验证了该技术路线的可行性及可操作性，对于口腔颅颌面外科软组织及颅底肿瘤等高难度复杂手术，可以提高手术精度，降低手术风险，并且降低外科医生的工作强度，具有很大的临床应用意义。