摘要
本文脑电图的自动分析已经成为生物医学电子学领域中的重要研究课题,目前所采用的分析方法主要有:脉冲响应法、幅度直方图、功率谱分析、相干函数与相关函数分析、压缩谱阵法和脑地形图分析等。但是专门讨论脑电波形识别的研究文献还不多见,现有的方法无论采用句法识别还是统计识别都比较复杂,而且没有考虑时间问题。我们在本文中提出了一种简单实用的方法,它对于癫痫病人脑电图中常见的两种病理波形(棘波和尖波)的识别非常有效。
Automatic analysis of EEG has become an important issue in the field of biome- bical electronics.To date,the main methods for analysis include pulse response,amp- litude histogram,spectral analysis,coherent function and crrelated function,compressed spectral array,and brain topography etc.Yet the soecialized papers on this subject are not appeared constantly.The methods now available,which use syntactic or statistic recognition,are too complicated.what is more,the time problem has not been con- sidered.In this paper,we propose a simple and practical method.It is very efficient to recognize two kinds of common pathologic EEG Waves of epiteptic patients.
关键词
脑电图
波形识别
棘波
尖波
electroencephalogram
wave recognition
spike wave
sharp wave
feature extraction
