Zschocke S, Rettig T, Rohr W
EEG EMG Z Elektroenzephalogr Elektromyogr Verwandte Geb. 1983 Jun;14(2):74-8.
This study demonstrates the possibilities of continuous frequency analysis supporting the evaluation of long lasting EEG recordings which present problems in analysing large amount of data. Avoiding some disadvantages of the usual presentation of power spectra by compressed spectral array compression of the spectral data will be obtained in a more suitable way by sequential plots of the data on the abscissa as time axis. The spectra will be separated in only a few number of frequency bands. A special conversion procedure (linear multiplication of the data within each frequency band by logarithmic calculated factors) provides for the essential enhancement of low spectral power in the upper EEG frequency domain preserving the full dynamic of spectral variations. The efficiency of this spectral trend analysis is shown by two examples: (1) In the study of the variations of cerebral excitability in a patient with epileptic seizures depending on the waking-sleeping-cycle (Fig. 3), and (2) in the documentation of the EEG dynamics in the course of intensive care monitoring. Fig. 5 shows the development of a cerebral autorhythmicity representing reintegration of cerebral function in a patient with severe head injury.
本研究证明了连续频率分析在支持对长时间脑电图记录进行评估方面的可能性，这些记录在分析大量数据时存在问题。通过将频谱数据按时间轴顺序绘制在横坐标上，能以更合适的方式获得频谱数据的压缩，从而避免了常规功率谱呈现方式的一些缺点。频谱将仅被分隔成少数几个频段。一种特殊的转换程序（在每个频段内将数据与对数计算因子进行线性相乘）可在保持频谱变化全动态范围的同时，显著增强脑电图高频域中的低频谱功率。这种频谱趋势分析的有效性通过两个例子得以展示：（1）在研究癫痫发作患者的大脑兴奋性变化与清醒 - 睡眠周期的关系时（图3），以及（2）在重症监护监测过程中记录脑电图动态变化时。图5展示了一名重度颅脑损伤患者大脑自律性的发展情况，这代表了大脑功能的重新整合。
