| 基于小波变换并结合神经网络的癫痫发作预报 |
| |
| 引用本文： | 林相波,邱天爽,李小兵,王静. 基于小波变换并结合神经网络的癫痫发作预报[J]. 中国生物医学工程学报, 2005, 24(5): 535-540 |
| |
| 作者姓名： | 林相波 邱天爽 李小兵 王静 |
| |
| 作者单位： | 1. 大连理工大学电子与信息工程学院,大连,116024
2. 北京天坛医院,北京,100101 |
| |
| 基金项目： | 国家自然科学基金资助项目（60172072,30170259,30570475）;辽宁省科学技术基金资助项目（2001101057）. |
| |
| 摘 要： | 采用信号处理的方法分析脑电图以实现自动预报癫痫发作是该领域一个难题，至今进展不够显著。本研究将小波变换用于脑电信号的预处理，并与递归神经网络RNN相结合预测癫痫发作。通过比较三种不同的预处理方法，发现在小波变换域利用脑电信号α节律的能量谱可以实现发作预报，而进一步提取包络并作非线性变换可以有效地提高RNN的预报性能。
|
| 关 键 词： | 癫痫 EEG 递归神经网络 小波变换 |
| 文章编号： | 0258-8021(2005)05-0535-06 |
| 收稿时间： | 2003-10-27 |
| 修稿时间： | 2004-03-25 |
The Prediction of Epileptic Seizures Based on the Wavelet Transform Combined With the Neural Network |
| |
| LIN Xiang-Bo,QIU Tian-Shuang,LI Xiao-Bing,WANG Jing. The Prediction of Epileptic Seizures Based on the Wavelet Transform Combined With the Neural Network[J]. Chinese Journal of Biomedical Engineering, 2005, 24(5): 535-540 |
| |
| Authors: | LIN Xiang-Bo QIU Tian-Shuang LI Xiao-Bing WANG Jing |
| |
| Abstract: | Automatic predicting the onset of epileptic seizure by signal processing on EEG is an important issue and difficult biomedical problem as well. Wavelet transform for signal preprocessing and recurrent neural network for feature recognition were used in this work to predict the onset of epileptic seizures. It was concluded by comparing three different preprocessing methods that the power spectra of alpha rhythm in wavelet transform domain could be used to predict epileptic seizures. The extraction of its envelope and nonlinear transform to the envelope provided further improvement of the performance of RNN effectively. |
| |
| Keywords: | epilepsy EEG recurrent neural network wavelet transform |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
