论文部分内容阅读
在对小波神经网络及其算法研究的基础上，提出了一种对脑电信号压缩表达和痫样脑电棘波识别的新方法。实验结果显示，小波网络在大量压缩数据的同时，能够较好的恢复原有信号。另外，在脑电信号的时频谱等高线图上，得到了易于自动识别的棘波和棘慢复合波特征，说明此方法在电生理信号处理和时频分析方面有着光明的应用前景。
Based on the research of wavelet neural network and its algorithm, a new method for compressive expression of EEG signals and identification of epileptiform EEG was proposed. Experimental results show that the wavelet network can compress the data a lot and recover the original signal well. In addition, the spikes and spike-slow complex signals that are easy to be automatically identified are obtained on the time-frequency contour map of EEG signals, indicating that this method has a bright application prospect in electrophysiological signal processing and time-frequency analysis.