摘要
癫痫和脑部肿瘤的神经外科手术所面临的一大难题是如何在切除病灶的同时不破坏正常脑功能区。因为语言功能区在大脑皮层中的分布复杂,个体差异大,在手术规划时确定关键语言功能区显得尤为重要。为此,本文采用一般线性模型(GLM),分析病人在执行语言任务时的大脑皮层脑电(ECoG)信号特征,得到癫痫病人在Beta(8~13Hz)和High gamma(60~90Hz)频段不同语言任务阶段的脑区参与情况。利用ECoG信号的高时间分辨率,区分了参与语言任务的运动区、语言准备区等。并将结果通过三维重建显示在病人个体脑结构上,提供了更加直观的语言功能在大脑皮层上的定位结果。通过与采用相同测试任务的功能性磁共振成像(fMRI)定位结果进行比较,本文中所采用的方法在准确定位语言功能区的同时,提供了时间分辨率更高(300ms^500ms)的定位结果。
Language function mapping has been one of the major concerns for epilepsy surgeries due to individual diversity and the complicated distribution over the cortex. The general linear model （GLM） was used here to analyze the power change of electrocorticograpic （ECoG） signals in the Beta （8 13 Hz） and High gamma （60 90 Hz） bands to obtain dynamic functional mapping for different stages during a single character word reading task. The high temporal resolution of ECoG can distinguish the motor cortex and related language preparing brain area. The mapping results were presented in a custom 3D visualization platform for individual brain structures to show more accurate and intuitive activation of the speech areas. Comparison with functional magnetic resonance imaging （fMRI） demonstrates that this method provides both reliable localization and temporal processing of language functional areas with high gamma responses using ECoG signals with high temporal resolution （300- 40（？ ms）.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2013年第9期1334-1338,共5页
Journal of Tsinghua University(Science and Technology)
基金
国家"九七三"重点基础研究项目(2012CB720704)
国家自然科学基金资助项目(61071003
90820304)
