Balestra V, Balzani P, Cabri M, Padovan C, Sannita W G
Boll Soc Ital Biol Sper. 1981 Sep 15;57(17):1823-9.
Data from power analysis of the human EEG were processed statistically, and an area within the power spectrum was identified, limited by the abscissa and by an exponential curve, the parameters of which could be defined by a moving linear regression after logarithmic transform of the spectral data. The EEG signal of normal subjects and epileptics with borderline EEG tracings was studied before and after administration of neuroactive compounds. The exponential curve was found to be negative in all subjects; its parameters were symmetrical and were not modified by the drugs administered, though these were active on the spectral profile. The area limited by the curve was used as a reference for the standardization of the power values with respect to a constant to the signal. The method allowed a better definition of the spectral profile than it is obtained by normalization vs. the total power value.
对人类脑电图功率分析的数据进行了统计处理,并在功率谱内确定了一个区域,该区域由横坐标和一条指数曲线限定,在对频谱数据进行对数变换后,通过移动线性回归可以定义该指数曲线的参数。在给予神经活性化合物之前和之后,对正常受试者和脑电图记录处于临界状态的癫痫患者的脑电图信号进行了研究。发现所有受试者的指数曲线均为负;其参数是对称的,并且不受所给药的影响,尽管这些药物对频谱特征有作用。曲线所限定的区域被用作相对于信号常数对功率值进行标准化的参考。该方法比通过相对于总功率值进行归一化能更好地定义频谱特征。