Hierarchical internal representation of spectral features in deep convolutional networks trained for EEG decoding

Recently, there is increasing interest and research on the interpretability of machine learning models, for example how they transform and internally represent EEG signals in Brain-Computer Interface (BCI) applications. This can help to understand the limits of the model and how it may be improved, in addition to possibly provide insight about the data itself. Schirrmeister et al. (2017) have recently reported promising results for EEG decoding with deep convolutional neural networks (ConvNets) trained in an end-to-end manner and, with a causal visualization approach, showed that they learn to use spectral amplitude changes in the input. In this study, we investigate how ConvNets represent spectral features through the sequence of intermediate stages of the network. We show higher sensitivity to EEG phase features at earlier stages and higher sensitivity to EEG amplitude features at later stages. Intriguingly, we observed a specialization of individual stages of the network to the classical EEG frequency bands alpha, beta, and high gamma. Furthermore, we find first evidence that particularly in the last convolutional layer, the network learns to detect more complex oscillatory patterns beyond spectral phase and amplitude, reminiscent of the representation of complex visual features in later layers of ConvNets in computer vision tasks. Our findings thus provide insights into how ConvNets hierarchically represent spectral EEG features in their intermediate layers and suggest that ConvNets can exploit and might help to better understand the compositional structure of EEG time series.

[1]  Joelle Pineau,et al.  Learning Robust Features using Deep Learning for Automatic Seizure Detection , 2016, MLHC.

[2]  Klaus-Robert Müller,et al.  Interpretable deep neural networks for single-trial EEG classification , 2016, Journal of Neuroscience Methods.

[3]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[4]  Trevor Darrell,et al.  Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[5]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[6]  Sepp Hochreiter,et al.  Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs) , 2015, ICLR.

[7]  Mariusz Bojarski,et al.  VisualBackProp: visualizing CNNs for autonomous driving , 2016, ArXiv.

[8]  A. Aertsen,et al.  The role of ECoG magnitude and phase in decoding position, velocity, and acceleration during continuous motor behavior , 2013, Front. Neurosci..

[9]  Klaus-Robert Müller,et al.  PatternNet and PatternLRP - Improving the interpretability of neural networks , 2017, ArXiv.

[10]  Wolfram Burgard,et al.  Deep learning with convolutional neural networks for EEG decoding and visualization , 2017, Human brain mapping.

[11]  Tara N. Sainath,et al.  Deep convolutional neural networks for LVCSR , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[12]  Yann LeCun,et al.  Convolutional networks and applications in vision , 2010, Proceedings of 2010 IEEE International Symposium on Circuits and Systems.

[13]  Rob Fergus,et al.  Visualizing and Understanding Convolutional Networks , 2013, ECCV.

[14]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[15]  Gerald Penn,et al.  Convolutional Neural Networks for Speech Recognition , 2014, IEEE/ACM Transactions on Audio, Speech, and Language Processing.

[16]  Andrew Zisserman,et al.  Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps , 2013, ICLR.

[17]  K.-R. Muller,et al.  Optimizing Spatial filters for Robust EEG Single-Trial Analysis , 2008, IEEE Signal Processing Magazine.

[18]  Stefan Haufe,et al.  On the interpretation of weight vectors of linear models in multivariate neuroimaging , 2014, NeuroImage.

[19]  Heesung Kwon,et al.  Single-trial EEG RSVP classification using convolutional neural networks , 2016, Defense + Security.