Behrens Alexander, Grimm Jonathan, Gross Sebastian, Aach Til
Faculty of Electrical Engineering and Information Technology, Institute of Imaging, & Computer Vision, RWTH Aachen University, Aachen 52062, Germany.
Annu Int Conf IEEE Eng Med Biol Soc. 2011;2011:5376-9. doi: 10.1109/IEMBS.2011.6091330.
The usage of video endoscopes in cystoscopic interventions of the urinary bladder impedes an intuitive navigation. Although image-based solutions such as panorama images can provide extended views of the surgical field, a real-time 3-D navigation is not supported. Furthermore, the integration of common tracking systems in ambulant clinics is often hindered due to low usability and high costs. Thus, we discuss in this paper a first low-cost inertial navigation system. Our evaluation results show that in spite of lower sensor accuracies, mean errors between < 1° and 4° are achieved for solid angles. Using endoscopes with different view angles we apply an extended endoscope model for an adaptive displacement correction. Furthermore, we implement a first guided navigation tool for tumor re-identification in real-time.
视频内窥镜在膀胱膀胱镜检查干预中的使用妨碍了直观导航。尽管基于图像的解决方案（如全景图像）可以提供手术视野的扩展视图，但不支持实时三维导航。此外，由于可用性低和成本高，移动诊所中常见跟踪系统的集成常常受到阻碍。因此，我们在本文中讨论了一种首个低成本惯性导航系统。我们的评估结果表明，尽管传感器精度较低，但立体角的平均误差在<1°至4°之间。使用具有不同视角的内窥镜，我们应用扩展内窥镜模型进行自适应位移校正。此外，我们还实现了首个用于实时肿瘤重新识别的引导导航工具。
Zotero 插件