Comparison of imaging modalities and source-localization algorithms in locating the induced activity during deep brain stimulation of the STN.
Comparison of imaging modalities and source-localization algorithms in locating the induced activity during deep brain stimulation of the STN.
复制标题
STN 深部脑刺激过程中定位诱发活动的成像方式和源定位算法的比较
DOI：
10.1109/embc.2016.7590651
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
M. Muthuraman
中科院分区：
文献类型：
--
作者：
K. G. Mideksa;A: Singh;N. Hoogenboom;H. Hellriegel;H. Krause;A. Schnitzler;G. Deuschl;J. Raethjen;G. Schmidt;M. Muthuraman
帕金森氏病(PD)患者最常用的治疗方法之一是丘脑底核(STN)的深部脑刺激(DBS)。寻找DBS电极放置的最佳靶区已成为当前研究的热点之一。在这项研究中，第一个目的是研究不同的源分析技术在探测位于皮质下水平的深部源并使用关于源位置的先验信息，即STN来验证它的能力。其次，我们的目标是研究脑电或脑磁图是否最适合绘制DBS诱导的大脑活动图。为此，使用脑电和脑磁图的同步测量来记录DBS诱发的电磁场和电势。采用边界元方法(BEM)求解正演问题。然后使用偶极(移动、旋转和固定MUSIC)和电流密度重建(CDR)(最小范数和sLORETA)方法估计DBS电极的位置。偶极子方法的震源定位结果表明，固定MUSIC算法对深部震源的定位效果最好，而运动偶极子算法不仅能检测感兴趣区域，还能检测受刺激STN影响的邻近区域。CDR方法的结果验证了sLORETA相对于最小范数检测STN的能力。此外，使用脑电模式的来源定位结果优于脑磁图，因为它定位了DBS在STN中的诱导活动。
One of the most commonly used therapy to treat patients with Parkinson's disease (PD) is deep brain stimulation (DBS) of the subthalamic nucleus (STN). Identifying the most optimal target area for the placement of the DBS electrodes have become one of the intensive research area. In this study, the first aim is to investigate the capabilities of different source-analysis techniques in detecting deep sources located at the sub-cortical level and validating it using the a-priori information about the location of the source, that is, the STN. Secondly, we aim at an investigation of whether EEG or MEG is best suited in mapping the DBS-induced brain activity. To do this, simultaneous EEG and MEG measurement were used to record the DBS-induced electromagnetic potentials and fields. The boundary-element method (BEM) have been used to solve the forward problem. The position of the DBS electrodes was then estimated using the dipole (moving, rotating, and fixed MUSIC), and current-density-reconstruction (CDR) (minimum-norm and sLORETA) approaches. The source-localization results from the dipole approaches demonstrated that the fixed MUSIC algorithm best localizes deep focal sources, whereas the moving dipole detects not only the region of interest but also neighboring regions that are affected by stimulating the STN. The results from the CDR approaches validated the capability of sLORETA in detecting the STN compared to minimum-norm. Moreover, the source-localization results using the EEG modality outperformed that of the MEG by locating the DBS-induced activity in the STN.