三维影像实时导航精准肝癌消融技术研究

Liver cancer is one of the four most common cancers with high incidence and mortality in China. Three-dimensional image guided liver cancer ablation is a major method to treat liver cancer. However, few effective methods are available to design ablation plan, which are essential for cases where multiple targets are required in order to completely ablate large hepatic tumors of irregular shape.Accurate target puncture is the premise of accurate ablation. The puncture accuracy can be improved using optical surgical navigation, but it is interfered by organ deformation induced by the factors, such as breath, heartbeat, and manual puncture introduces great error. This challenge is still not solved. For these reasons, this study focuses on how to design intelligent ablation plan. An optimized mathematical model of ablation will be constructed, which aims to get preoperative three-dimensional images to compute the parameters, such as target number and positions, surgical routes, ablating and dose of each target. A new type of fiducial marker will be designed and then an automatic technique of intraoperative registration between image space and patient space will be studied to realize registration in real time in the process of tracking, which can improve the accuracy and speed of puncture. The optimal timeslot will be accurately computed to effectively restrain the influences, such as breath, heartbeat. The control algorithm will be studied to control the surgical robot to perform puncture in optimal timeslots. This procedure does not require manual intervention. Therefore, the multiple target puncture will be accurately and rapidly performed. The ablation devices will be controlled to accurately, conformally, and appropriately ablate according to the ablating time and dose of each target in the ablation plan.

肝癌是我国发病率和死亡率最高的四种癌症之一，三维影像引导肝癌消融是肝癌手术的重要手段，但缺乏良好的消融方案设计方法，对于肿瘤形状不规则的大肝癌等需多靶点消融情况尤为重要。精准靶点穿刺是精准消融的前提。虽然采用光学手术导航可提高穿刺精度，但术中呼吸和心跳等因素导致器官形变严重影响穿刺精度，人工穿刺误差大，尚无有效方法解决此难题。为此，本项目重点研究智能化消融方案设计方法，建立最优化消融数学模型，通过术前三维影像获取消融靶点数量、位置、手术路径及各靶点消融时间和剂量。设计新型标记点，研究术中图像空间与实际空间的自动注册技术，在跟踪定位时实现实时注册，提高注册精度与速度。精确计算最佳穿刺时隙，有效抑制呼吸心跳等因素影响。研究机器人控制算法控制手术机器人在最佳时隙内进行穿刺，穿刺过程无需人工介入，实现精准、快速多靶点穿刺。根据消融方案中各靶点消融时间和剂量，控制消融设备进行精准、适形、适量消融。

目前肝癌消融手术主要靠医生根据术前影像资料，在超声或CT图像引导下凭借其主观经验进行。超声实时图像分辨率较低，导致穿刺精度较低；而CT图像分辨率较高，但不具备实时性，需依靠医生丰富的临床操作经验，多次穿刺才能准确到达靶区，从而增加病人组织损伤以及加大医生工作强度。因此，阻碍精准肝癌消融技术发展的关键是消融方案制定和对消融靶点的精准定位与穿刺问题。对此，本项目对肝癌精准消融技术进行了深入研究，主要完成了以下内容：1、构建了肝癌精准消融软硬件系统，其主要由光学定位系统、六自由度机械臂、消融针以及相关软件算法等构成；2、完成了有限元肝脏消融模型的研究，实现最佳肝癌消融方案的制定；3、完成了图像空间标记点提取算法、实际空间标记点提取算法和手术空间注册算法以及机器人手眼标定算法，实现系统准确标定和靶点精准定位；4、完成了基于三维空间几何约束的穿刺机器人靶区定位算法以及最佳呼吸穿刺时隙下的机器人跟随与穿刺策略，以抑制呼吸运动对穿刺精度的影响。本项目取得的重要结果和关键数据包括：1、光学定位系统图像采集与传输帧率达60帧/秒、标记点定位精度达0.15毫米，满足精准肝癌消融系统对实时性和准确性的要求；2、手术空间注册精度在0.8毫米以内，注册时间小于8毫秒，为医生提供精准实时的手术导航信息；3、手术机器人标定平均误差低于2毫米；4、呼吸运动仿真实验装置输出的信号单个周期约6.5秒，最大运动位移达35毫米，符合人体呼吸运动的特征；5、呼吸运动下手术机器人靶区穿刺误差在2.5毫米以下，满足临床肝癌消融的要求。总之，本项目研究的肝癌精准消融技术，具有消融方案制定智能化、手术空间注册自动化、消融针定位精度高等特点，为光学导航技术与手术机器人技术结合的精准肝癌消融技术的进一步推广和应用打下了良好的基础。本项目的主要成果有SCI论文8篇，中文核心期刊论文1篇，发明专利11项，实用新型专利2项，软件著作1项，共培养研究生16人。