论文部分内容阅读
◆Abstract：Sleep is a very important life process， but until now， very little was known about it. Based on MATLAB simulation system， this paper mainly studies the application of wavelet transform in EEG signal processing， and the frequency signals of each layer obtained by wavelet decomposition are extracted.
◆Keywords：Brain electrical signal;The wavelet transform;Denoising refactoring
1 The Introduction
EEG is a basic physiological signal of human body， which contains rich physiological. The purpose of this paper is to analyze and process the EEG signals obtained by the EEG acquisition instrument by using wavelet transform and other methods.
2 EEG signal analysis
According to their frequency and amplitude， brain waves can be divided into four basic types： delta， theta， alpha， and beta. The origin and function of the four waveforms are also different. The EEG data used in this paper are acquired by using the EEG acquisition system. EEG collection uses 16 channels with a sampling frequency of 256Hz.
3 The wavelet transform
Brain signal preprocessing can filter out the power frequency interference and some noise in the collected EEG signal， so the signal-noise separation plays an important role in the further processing of the signal： low entropy， de-correlation， multi-resolution and flexibility of base selection.
4 Feature extraction
Because of the obvious nonlinear characteristics of EEG signals， nonlinear dynamics has been applied to the analysis of EEG signals more and more widely in recent years， including complexity and other analysis methods. In order to extract multiple sleep features from sleep EEG， this paper uses sample entropy nonlinear dynamics method to extract sleep features from sleep EEG.
5 Conclusion
In this paper， the wavelet threshold method is used to denoise the EEG signal， and then the sleep EEG after denoising is calculated as the characteristic of sleep stage. The conclusions are as follows ： EEG signals belong to non-stationary random signals， and wavelet analysis method can directly observe some frequency components of signals or extract useful characteristic signals.
Reference
[1]RECHTSCHAFFEN A，KALES A.A manual of standardi-zed terminology，techniques and scoring system for sleep stage of human subjects[M].Washington D.C.，USA：Government Printing Office，1968.
◆Keywords：Brain electrical signal;The wavelet transform;Denoising refactoring
1 The Introduction
EEG is a basic physiological signal of human body， which contains rich physiological. The purpose of this paper is to analyze and process the EEG signals obtained by the EEG acquisition instrument by using wavelet transform and other methods.
2 EEG signal analysis
According to their frequency and amplitude， brain waves can be divided into four basic types： delta， theta， alpha， and beta. The origin and function of the four waveforms are also different. The EEG data used in this paper are acquired by using the EEG acquisition system. EEG collection uses 16 channels with a sampling frequency of 256Hz.
3 The wavelet transform
Brain signal preprocessing can filter out the power frequency interference and some noise in the collected EEG signal， so the signal-noise separation plays an important role in the further processing of the signal： low entropy， de-correlation， multi-resolution and flexibility of base selection.
4 Feature extraction
Because of the obvious nonlinear characteristics of EEG signals， nonlinear dynamics has been applied to the analysis of EEG signals more and more widely in recent years， including complexity and other analysis methods. In order to extract multiple sleep features from sleep EEG， this paper uses sample entropy nonlinear dynamics method to extract sleep features from sleep EEG.
5 Conclusion
In this paper， the wavelet threshold method is used to denoise the EEG signal， and then the sleep EEG after denoising is calculated as the characteristic of sleep stage. The conclusions are as follows ： EEG signals belong to non-stationary random signals， and wavelet analysis method can directly observe some frequency components of signals or extract useful characteristic signals.
Reference
[1]RECHTSCHAFFEN A，KALES A.A manual of standardi-zed terminology，techniques and scoring system for sleep stage of human subjects[M].Washington D.C.，USA：Government Printing Office，1968.