Electroencephalogram signal feature extraction and classification method of combining DAE (denoising auto encoder) and CNN (convolutional neural network)

Abstract

The invention discloses an electroencephalogram signal feature extraction and classification method of combining a denoising auto encoder (DAE) and a convolutional neural network (CNN). The method includes the steps of: collecting electroencephalogram data through an electroencephalogram signal collection instrument; carrying out preprocessing of abnormal-sample removing, mean removing, signal filtering and the like on the collected data; using the auto encoder, into which a noise coefficient is added, and electroencephalogram signals for training; using hidden layers of the denoising auto encoder as feature data output; then converting obtained feature data into an image-like format; utilizing the convolutional neural network for classification; and finally, utilizing a test data set to carry out a performance test on the trained network. Compared with other traditional methods, the method of the invention can obtain a higher classification accuracy rate and higher robustness.

Description

technical field [0001] The invention belongs to a method for feature extraction and classification of electroencephalogram signals, in particular to a method for feature extraction and classification of electroencephalogram signals combined with a noise reduction automatic encoding machine and a convolutional neural network. Background technique [0002] Brain-computer interface (BCI) is independent of peripheral nerve tissue and directly establishes a communication channel with external devices. Since it was first proposed, it has become a research hotspot in the fields of brain science and cognitive science. In a brain-computer interface system, signal recognition usually includes three parts: preprocessing, feature extraction, and classification. [0003] In the traditional method, in terms of preprocessing: using wavelet transform, ICA processing, spatial filtering and other methods, the present invention carries out signal preprocessing in three steps with reference to ...

AI Technical Summary

Problems solved by technology

In terms of feature extraction: common spatial pattern (CSP) is used to extract features of motor imagery, but time domain analysis is too expensive and requires a high number of EEG channels; autoregressive model method (AR) is used for prediction, but the AR model is suitable for Single-channel data has limitations for complex and high-dimensional EEG signals, and the classification accuracy is not high

In terms of classification methods: use linear discriminant analysis (LDA), but LDA is suitable for linear samples, not suitable for the nonlinear EEG data mentioned in this article; use support vector machine (SVM), SVM can better solve complex nonlinear data, but as a supervised network, labels are required in the training and testing process, and parameter adjustment is complicated

Images