论文部分内容阅读
脑电(EEG)癫痫波的自动检测与分类在临床医学上具有重要意义。针对EEG信号的非平稳特点,本文提出了一种基于经验模式分解(EMD)和支持向量机(SVM)的EEG分类方法。首先利用EMD将EEG信号分成多个经验模式分量,然后提取有效特征,最后用SVM对EEG信号进行分类。结果表明,该方法对癫痫发作间歇期和发作期EEG的分类效果比较理想,识别率达到99%。
EEG Seizure wave automatic detection and classification of clinical medicine is of great significance. In view of the non-stationary characteristics of EEG signals, this paper proposes an EEG classification method based on Empirical Mode Decomposition (EMD) and Support Vector Machine (SVM). Firstly, the EEG signal is divided into multiple empirical mode components by using EMD, and then the valid features are extracted. Finally, the EEG signals are classified by SVM. The results show that this method is ideal for the classification of seizure interval and episode EEG, the recognition rate of 99%.