Functional localization and restoration of language areas based on the cerebral oscillatory changes

基于神经磁振荡变化的语言优势与WADA测试的符合率为85%。我们的方法与模拟标测的定位误差为6.3 mm，这表明我们的方法是一种可行的无创方法。滑动时间窗分析显示，θ振荡反映连续加工，α去同步化反映感觉运动功能，β/低γ去同步化反映负运动区和运动语言区的活动。皮层脑电图清楚地检测到γ振荡。使用支持向量机对三种类型的单音节发音过程中的皮层脑电进行解码。它的译码精度与发声同相达到60%-70%的峰值。

Language dominance based on the neuromagnetic oscillatory changes was coincident in 85% with that by the Wada test. Difference in localization between our method and simulation mapping was 6.3mm, which indicates that our method is a feasible noninvasive method. Sliding time window analyses revealed that θ oscillation reflects serial processing, α desynchronization reflects sensorimotor function, and β/low γ desynchronization reflects the activities in the negative motor area and motor language area. Electrocorticograms clearly detected γ oscillation. Electrocorticograms during three types of single syllable phonation were decoded using a support vector machine. Its decoding accuracy peaked at 60-70% in phase with phonation.