Spontaneous EEG Classification Using Complex Valued Neural Network

Identification of spontaneous brain activity using the electroencephalography (EEG) requires information of the frequency spectrum and the spatial distribution. The complex valued neural network (CVNN) which uses complex weights and inputs has been shown higher performance for periodic data analysis, since spectrum information is represented by complex numbers. In spontaneous EEG analysis, the phase information depends on the onset of the recording, thus it is not informative. However, the conventional CVNN is not able to remove the phase information and extract amplitude spectrum efficiently. In this paper, we introduce two activation functions for CVNN to extract the amplitude spectrum directly, and classify spontaneous EEG. Our experimental results showed that the proposed method is higher classification performance than the conventional CVNN, and comparable to the convolutional neural network (CNN). Furthermore, the proposed method showed high performance when the number of hidden units is small.

[1]  Leontios J. Hadjileontiadis,et al.  A Novel Emotion Elicitation Index Using Frontal Brain Asymmetry for Enhanced EEG-Based Emotion Recognition , 2011, IEEE Transactions on Information Technology in Biomedicine.

[2]  Tülay Adali,et al.  Approximation by Fully Complex Multilayer Perceptrons , 2003, Neural Computation.

[3]  Tingwen Huang,et al.  Learning of Phase-Amplitude-Type Complex-Valued Neural Networks with Application to Signal Coherence , 2017, ICONIP.

[4]  N. Birbaumer,et al.  BCI2000: a general-purpose brain-computer interface (BCI) system , 2004, IEEE Transactions on Biomedical Engineering.

[5]  Akira Hirose,et al.  Proposal of Complex-Valued Convolutional Neural Networks for Similar Land-Shape Discovery in Interferometric Synthetic Aperture Radar , 2018, ICONIP.

[6]  Wenming Zheng,et al.  Multichannel EEG-Based Emotion Recognition via Group Sparse Canonical Correlation Analysis , 2017, IEEE Transactions on Cognitive and Developmental Systems.

[7]  Jeffrey M. Hausdorff,et al.  Physionet: Components of a New Research Resource for Complex Physiologic Signals". Circu-lation Vol , 2000 .

[8]  Haipeng Wang,et al.  Complex-Valued Convolutional Neural Network and Its Application in Polarimetric SAR Image Classification , 2017, IEEE Transactions on Geoscience and Remote Sensing.

[9]  Sandeep Subramanian,et al.  Deep Complex Networks , 2017, ICLR.

[10]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[11]  T. Nitta,et al.  A back-propagation algorithm for complex numbered neural networks , 1993, Proceedings of 1993 International Conference on Neural Networks (IJCNN-93-Nagoya, Japan).

[12]  Igor Aizenberg,et al.  Analysis of EEG Using Multilayer Neural Network with Multi-Valued Neurons , 2018, 2018 IEEE Second International Conference on Data Stream Mining & Processing (DSMP).

[13]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[14]  Junming Zhang,et al.  A New Method for Automatic Sleep Stage Classification , 2017, IEEE Transactions on Biomedical Circuits and Systems.

[15]  Musa Peker,et al.  A Novel Method for Automated Diagnosis of Epilepsy Using Complex-Valued Classifiers , 2016, IEEE Journal of Biomedical and Health Informatics.

[16]  Toby P. Breckon,et al.  On the Relative Contribution of Deep Convolutional Neural Networks for SSVEP-Based Bio-Signal Decoding in BCI Speller Applications , 2019, IEEE Transactions on Neural Systems and Rehabilitation Engineering.