Zelmann R, Beriault S, Marinho M M, Mok K, Hall J A, Guizard N, Haegelen C, Olivier A, Pike G B, Collins D L
McConnell Brain Imaging Center, Montreal Neurological Hospital and Institute, McGill University, 3801 University Street, Montreal, QC, H3A 2B4, Canada.
Neurology and Neurosurgery, Montreal Neurological Hospital and Institute, McGill University, 3801 University Street, Montreal, QC, H3A 2B4, Canada.
Int J Comput Assist Radiol Surg. 2015 Oct;10(10):1599-615. doi: 10.1007/s11548-015-1165-6. Epub 2015 Mar 26.
Intracranial electrodes are sometimes implanted in patients with refractory epilepsy to identify epileptic foci and propagation. Maximal recording of EEG activity from regions suspected of seizure generation is paramount. However, the location of individual contacts cannot be considered with current manual planning approaches. We propose and validate a procedure for optimizing intracranial electrode implantation planning that maximizes the recording volume, while constraining trajectories to safe paths.
Retrospective data from 20 patients with epilepsy that had electrodes implanted in the mesial temporal lobes were studied. Clinical imaging data (CT/A and T1w MRI) were automatically segmented to obtain targets and structures to avoid. These data were used as input to the optimization procedure. Each electrode was modeled to assess risk, while individual contacts were modeled to estimate their recording capability. Ordered lists of trajectories per target were obtained. Global optimization generated the best set of electrodes. The procedure was integrated into a neuronavigation system.
Trajectories planned automatically covered statistically significant larger target volumes than manual plans [Formula: see text]. Median volume coverage was [Formula: see text] for automatic plans versus [Formula: see text] for manual plans. Furthermore, automatic plans remained at statistically significant safer distance to vessels [Formula: see text] and sulci [Formula: see text]. Surgeon's scores of the optimized electrode sets indicated that 95% of the automatic trajectories would be likely considered for use in a clinical setting.
This study suggests that automatic electrode planning for epilepsy provides safe trajectories and increases the amount of information obtained from the intracranial investigation.
对于难治性癫痫患者，有时会植入颅内电极以识别癫痫病灶和传播途径。从疑似癫痫发作起源区域最大程度地记录脑电图（EEG）活动至关重要。然而，目前的手动规划方法无法考虑单个电极触点的位置。我们提出并验证了一种优化颅内电极植入规划的程序，该程序在将轨迹限制在安全路径的同时，能使记录体积最大化。
研究了20例在颞叶内侧植入电极的癫痫患者的回顾性数据。对临床影像数据（CT/A和T1加权磁共振成像）进行自动分割，以获取目标区域和需避开的结构。这些数据被用作优化程序的输入。对每个电极进行建模以评估风险，同时对单个电极触点进行建模以估计其记录能力。获得每个目标的有序轨迹列表。全局优化生成最佳电极组合。该程序被集成到神经导航系统中。
自动规划的轨迹在统计学上覆盖的目标体积比手动规划的大得多[公式：见原文]。自动规划的中位体积覆盖率为[公式：见原文]，而手动规划的为[公式：见原文]。此外，自动规划与血管[公式：见原文]和脑沟[公式：见原文]的距离在统计学上显著更安全。外科医生对优化电极组合的评分表明，95%的自动轨迹可能会被考虑用于临床。
本研究表明，癫痫的自动电极规划可提供安全的轨迹，并增加从颅内检查中获得的信息量。
