Goeller C J, Sinton C M
Research Department, CIBA-GEIGY Corporation, Summit, NJ 07901.
Comput Methods Programs Biomed. 1989 May;29(1):31-6. doi: 10.1016/0169-2607(89)90087-4.
A microcomputer-based sleep stage analysis system for laboratory use is described. Rapid spectral analysis of the electroencephalogram (EEG) is achieved through hardware; subsequent analysis, including determination of the relative power in user-defined EEG frequency bands is provided through software. A decision matrix, based on standard sleep state analysis criteria, enables these spectral data, in conjunction with integrated electromyogram (EMG) power, to determine sleep stages. The system has been configured here to quantify standard sleep stages, but the algorithm can also identify, and provide information about, EEG states which are not easily classified according to earlier criteria used for visual classification of polygraphic records. This low-cost, user-friendly system thus achieves flexible, quantitative analysis of EEG and EMG signals.
本文介绍了一种用于实验室的基于微型计算机的睡眠阶段分析系统。脑电图（EEG）的快速频谱分析通过硬件实现；后续分析，包括在用户定义的EEG频段中确定相对功率，通过软件提供。基于标准睡眠状态分析标准的决策矩阵，结合整合肌电图（EMG）功率，使这些频谱数据能够确定睡眠阶段。该系统在此配置为量化标准睡眠阶段，但该算法还可以识别并提供有关EEG状态的信息，这些状态根据早期用于多导睡眠图记录视觉分类的标准不易分类。因此，这个低成本、用户友好的系统实现了对EEG和EMG信号的灵活、定量分析。
