癫痫患者脑磁图的高频振荡量化算法研究

Quantification of high-frequency oscillations for epileptic MEG

目的:验证重心频率能否有效量化癫痫患者脑磁图中的高频振荡成分。方法:选择1例典型颞叶癫痫患者的脑磁图数据,采用短时傅里叶变换进行分解,用重心频率对高频振荡进行量化,并计算平均重心频率,比较左、右颞的重心频率差异。结果:在静息状态下,患侧的梯度计采集到的MEG信号在80～250Hz范围内即时重心频率和平均重心频率都比健侧高。结论:采用重心频率能够有效的量化信号的高频振荡,同时能够有效的区分出健侧和患侧信号之间的高频振荡差异。

Objective：To evaluate the validity of gravity frequency used for quantifying the high-frequency oscillations of epileptic MEG.Methods：A temporal lobe epileptic MEG data was decomposed by short time Fourier transforms.Then,the average gravity frequency was calculated.The difference of average gravity frequency between LT and RT gradient sensor was obtained.Results：The gravity frequency from frequency band 80-250 Hz of MEG from injured side was higher than that from uninjured side.Conclusion：The high-frequency oscillations for epileptic MEG can be quantified by gravity frequency.The HFOs from injured and uninjured side can be distinguished.