Jackson Peter, Simon Richard, Linte Cristian
Annu Int Conf IEEE Eng Med Biol Soc. 2020 Jul;2020:5081-5084. doi: 10.1109/EMBC44109.2020.9175270.
Most image-guided interventions rely on surgical tracking and image/model to patient registration to establish a spatial relationship between the patient and the pre- and intraprocedural images, by using surgical tracking and localization systems. In this work, we characterize the tracking, registration and navigation accuracy using two different surgical localization systems - the NDI Polaris Spectra optical tracking system and the NDI Aurora electromagnetic tracking system - in the context of an image-guided renal intervention, using a 3D printed life-size model of a patient-specific kidney phantom generated from a CT image. Our results reported a 0.05 mm fiducial localization error, 0.70 mm fiducial registration error, and 0.78 mm target registration error, and 0.63 mm overall navigation error using the optical tracking, and 0.12 mm fiducial localization error, 0.78 mm fiducial registration error, 0.93 mm target registration error and 0.89 mm overall navigation error using electromagnetic tracking. Additionally, our study also showed similarity between the overall navigation accuracy using optical (0.63 mm RMS error) or electromagnetic tracking (0.89 mm RMS error) and the overall navigation accuracy achieved using direct visualization of the surgical scene (0.68 mm and 1.06 RMS error respectively), which serves as a baseline control metric.
大多数图像引导介入手术依靠手术追踪以及图像/模型到患者的配准，通过使用手术追踪和定位系统来建立患者与术前和术中图像之间的空间关系。在这项工作中，我们在图像引导肾介入手术的背景下，使用由CT图像生成的患者特异性肾脏模型的3D打印实物大小模型，利用两种不同的手术定位系统——NDI Polaris Spectra光学追踪系统和NDI Aurora电磁追踪系统，对追踪、配准和导航精度进行了表征。我们的结果显示，使用光学追踪时，基准定位误差为0.05毫米，基准配准误差为0.70毫米，目标配准误差为0.78毫米，总体导航误差为0.63毫米；使用电磁追踪时，基准定位误差为0.12毫米，基准配准误差为0.78毫米，目标配准误差为0.93毫米，总体导航误差为0.89毫米。此外，我们的研究还表明，使用光学追踪（均方根误差为0.63毫米）或电磁追踪（均方根误差为0.89毫米）的总体导航精度与通过直接观察手术场景实现的总体导航精度（分别为0.68毫米和1.06均方根误差）之间具有相似性，后者作为基线控制指标。
