NEW METHODS TO IMPROVE SPATIAL RESOLUTION OF EEG

拟议的工作涉及标准方法的评价， 开发新的方法来分析多通道EEG数据， 物理学、工程学和计算机发展的结合， 在EEG中建立方法。 一个主要目标是改善 EEG空间分辨率，使头皮记录更准确 表示潜在来源的活动。 另一个目标是， 开发方法来量化EEG的空间模式， 各种临床和认知状态。 空间模式可以是 潜在振幅，拉普拉斯（电流源密度）估计， 相关性或相干性模式。 这种方法的独特之处 包括局部颅骨阻力实验和理论估计， 癫痫发作和发作间期现象的综合研究 患者，分析和人体头部的有限元模型， 同时记录猫的头皮和皮层EEG（64个通道）， 应用各种样条估计EEG中的表面拉普拉斯算子， 诱发电位，以及空间去卷积方法在 多通道头皮数据。 所提出的方法适用于 分布式和时变源。 因此，这项工作可望 直接适用于各种未来的脑电图和诱发 在许多地点进行了潜力研究。

The proposed work is concerned with the evaluation of standard methods and the development of new methods to analyze multichannel EEG data base on the integration of physics, engineering, and computer developments with established methods in EEG. A principal objective is to improve the spatial resolution of EEG so that scalp recordings more accurately represent the activity of underlying sources. Another goal is that of the development of methods to quantify spatial patterns of EEG associated with various clinical and cognitive states. The spatial patterns may be potential amplitudes, Laplacian (current source density) estimates, correlation, or coherency patterns. Unique features of this approach include experimental and theoretical estimates of local skull resistance, a comprehensive study if ictal and interictal phenomena in epilepsy patients, analytic and finite element models of the human head, simultaneous recordings of scalp and cortical EEG in cats (64 channels), application of various splines to estimate surface Laplacians in EEG and evoked potentials, and application of spatial deconvolution methods to multichannel scalp data. The proposed methods are applicable to distributed and time-varying sources. Thus, this work can be expected to be directly applicable to a wide variety of future EEG and evoked potentials studies carried out at numerous locations.