Shibasaki H, Nakamura M, Nishida S
Electroencephalogr Clin Neurophysiol. 1987 Feb;66(2):200-4. doi: 10.1016/0013-4694(87)90191-x.
A new method of wave form decomposition by using an evoked potential model was applied to the study of scalp topography of photic evoked potentials in 3 normal subjects. The parameters of the evoked potential model were composed of gain, natural frequency, damping ratio and time lag, and were determined in the frequency domain to minimize the square sums of the modeling error. By using the optimal model parameters, the wave form was decomposed into basic components. At least 4 components occurring within the initial 200 msec period after the stimulus onset were identified in all subjects and were shown to have different scalp distributions, suggesting different generator sources.
一种利用诱发电位模型进行波形分解的新方法被应用于对3名正常受试者光诱发电位头皮地形图的研究。诱发电位模型的参数由增益、固有频率、阻尼比和时间延迟组成,并在频域中确定,以使建模误差的平方和最小化。通过使用最优模型参数,将波形分解为基本成分。在所有受试者中均识别出至少4个在刺激开始后最初200毫秒内出现的成分,且这些成分具有不同的头皮分布,提示其产生源不同。