Localization of implanted EEG electrodes in a virtual-reality environment.

Localization of implanted EEG electrodes in a virtual-reality environment.
复制标题

DOI:
10.3109/10929080109146090
复制
发表时间:
2001-01-01
期刊:
Computer aided surgery : official journal of the International Society for Computer Aided Surgery
影响因子:
--
通讯作者:
Hoekema, R
Hoekema, R
中科院分区:
其他
文献类型:
--
作者:
Noordmans, H J;van Rijen, P C;Hoekema, R

文献摘要

被引文献

相似文献

在癫痫手术程序的计划中,很大一部分患者植入了颅内电极。个体电极接触相对于大脑皮层的精确定位是必要的。由于在CT扫描中以逐片方式手动跟踪脑电图电极是繁琐和主观的,因此本研究的目标是开发一种更容易和更准确的方法来定位植入的脑电图电极。在本文中,我们提出了我们的解决方案,以虚拟现实环境的形式与交互工具,以协助临床医生的脑电图定位。在高质量和快速体积渲染器的帮助下,创建了患者颅骨内部的视图,以获得与皮质结构相关的电极的概述。深度电极、栅格电极和簧片电极使用不同的方法进行半交互表征。对于深度电极,触点(在CT扫描中不可见)是通过测量触点和电极末端从三维(3D)线跟踪器产生的中心轴之间的理论距离得出的。对于网格电极,触点在CT上是可见的,因此3D视图仅用于查找触点并解决网格与其他网格、尾线或骨脊的重叠。对于簧片电极,在这种情况下同样不可见的触点是根据拟合到引线标记位置的线模型计算的。在让用户将人造球体放置在引线标记和导线上之后,b样条被安装到球体的中心以估计触点的位置。通过将该方法应用于7例患者的CT扫描来评估该方法。该方法似乎是普遍适用的(即使是交叉电极或有间隙的电极也能被正确表征),并且通过3D视图和切片的显示非常好,手动放置球体的效果与半自动放置一样好。计算机实验表明,脑电触点位置的最终定位误差可以估计在一个体素的维数的顺序上。
In the planning of epilepsy surgery procedures, intracranial electrodes are implanted in a significant fraction of the patients. Accurate localization of the individual electrode contacts with respect to the brain cortex is imperative. Because the manual tracking of an EEG electrode in a CT scan in a slice-by-slice fashion is cumbersome and subjective, the goal of this study was to develop an easier and more accurate way to localize implanted EEG electrodes. In this paper, we present our solution in the form of a virtual-reality environment with interactive tools to assist the clinician with EEG localization. With the help of a high-quality and fast volume renderer, a view is created of the inside of the patient's skull to obtain an overview of the electrodes in relation to the cortical structures. Depth, grid, and reed electrodes are characterized semi-interactively using different methods. For depth electrodes, the contacts (which are not visible in the CT scan) are derived by measuring off the theoretical distance between the contact and the end of the electrode from the central axis produced by a three-dimensional (3D) line tracker. For grid electrodes, the contacts are visible in a CT, so the 3D view is merely used to find the contacts and to resolve the overlap of grids with other grids, tail wires, or bone ridges. For reed electrodes, the contacts, which are again not visible in this case, are calculated from a line model fitted to the positions of lead markers. After letting the user place artificial spheres on the lead markers and wire, a B-spline is fitted to the spheres' centers to estimate the positions of the contacts. The approach was evaluated by applying it to CT scans of seven patients. It appeared that the method is generally applicable (even crossing electrodes or electrodes with gaps were correctly characterized), and that the display via 3D views and slices is so good that manual placement of spheres performed as well as semi-automatic placement. From computer experiments, it appeared that the final localization error in the position of EEG contacts could be estimated to lie in the order of the dimensions of one voxel.