Fusaglia Matteo, Tinguely Pascale, Banz Vanessa, Weber Stefan, Lu Huanxiang
University of Bern, Bern, Switzerland
University Hospital of Bern, Bern, Switzerland.
Surg Innov. 2016 Aug;23(4):397-406. doi: 10.1177/1553350616637691. Epub 2016 Mar 10.
Background Patient-to-image registration is a core process of image-guided surgery (IGS) systems. We present a novel registration approach for application in laparoscopic liver surgery, which reconstructs in real time an intraoperative volume of the underlying intrahepatic vessels through an ultrasound (US) sweep process. Methods An existing IGS system for an open liver procedure was adapted, with suitable instrument tracking for laparoscopic equipment. Registration accuracy was evaluated on a realistic phantom by computing the target registration error (TRE) for 5 intrahepatic tumors. The registration work flow was evaluated by computing the time required for performing the registration. Additionally, a scheme for intraoperative accuracy assessment by visual overlay of the US image with preoperative image data was evaluated. Results The proposed registration method achieved an average TRE of 7.2 mm in the left lobe and 9.7 mm in the right lobe. The average time required for performing the registration was 12 minutes. A positive correlation was found between the intraoperative accuracy assessment and the obtained TREs. Conclusions The registration accuracy of the proposed method is adequate for laparoscopic intrahepatic tumor targeting. The presented approach is feasible and fast and may, therefore, not be disruptive to the current surgical work flow.
背景 患者到图像配准是图像引导手术（IGS）系统的核心过程。我们提出了一种新的配准方法，用于腹腔镜肝手术，该方法通过超声（US）扫描过程实时重建肝内血管的术中容积。方法 对现有的开放式肝脏手术IGS系统进行了改进，对腹腔镜设备进行了合适的器械跟踪。通过计算5个肝内肿瘤的目标配准误差（TRE），在逼真的模型上评估配准精度。通过计算执行配准所需的时间来评估配准工作流程。此外，还评估了一种通过将超声图像与术前图像数据进行视觉叠加来进行术中精度评估的方案。结果 所提出的配准方法在左叶的平均TRE为7.2毫米，在右叶为9.7毫米。执行配准所需的平均时间为12分钟。术中精度评估与获得的TRE之间存在正相关。结论 所提出方法的配准精度足以用于腹腔镜肝内肿瘤靶向。所提出的方法可行且快速，因此可能不会干扰当前的手术工作流程。
