A novel deep learning approach for classification of EEG motor imagery signals

By Yousef Rezaei Tabar, Ugur Halici. Journal of Neural Engineering, 2017

823 citations recorded by the source corpus.

Objective. Signal classification is an important issue in brain computer interface (BCI) systems. Deep learning approaches have been used successfully in many recent studies to learn features and classify different types of data. However, the number of studies that employ these approaches on BCI applications is very limited. In this study we aim to use deep learning methods to improve classification performance of EEG motor imagery signals. Approach. In this study we investigate convolutional neural networks (CNN) and stacked autoencoders (SAE) to classify EEG Motor Imagery signals. A new form of input is introduced to combine time, frequency and location information extracted from EEG signal and it is used in CNN having one 1D convolutional and one max-pooling layers. We also proposed a new deep network by combining CNN and SAE. In this network, the features that are extracted in CNN are classified through the deep network SAE. Main results. The classification performance obtained by the proposed method on BCI competition IV dataset 2b in terms of kappa value is 0.547. Our approach yields 9% improvement over the winner algorithm of the competition. Significance. Our results show that deep learning methods provide better classification performance compared to other state of art approaches. These methods can be applied successfully to BCI systems where the amount of data is large due to daily recording.

Subfields: Cognitive Neuroscience, Human-Computer Interaction

View Paper: https://www.semanticscholar.org/paper/6053679654c06315f39986535a343085b07b3efd

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Human-Computer Interaction

Citations in the corpus

TSception: Capturing Temporal Dynamics and Spatial Asymmetry From EEG for Emotion Recognition (2021, methodology): “…fusion to detect emotions from EEG. Recently, deep learning-based methods have shown promising results in the BCI domain, such as motor imagery clas-siﬁcation [15], [16], [17], [18], [19], emotion recognition [20], [21], [22], [23], [24], [25], and mental-task classiﬁcation [26], [27], [28].”
DADA: Deep Adversarial Data Augmentation for Extremely Low Data Regime Classification (2018, methodology): “Recent works [41, 42] have explored CNNs in classifying EEG signals.”
HS-CNN: a CNN with hybrid convolution scale for EEG motor imagery classification (2020, methodology): “For time-frequency feature extraction, there are many works that use short-time Fourier transform (STFT) [18, 30] or wavelets [19–21, 31] to extract the time-frequency features of EEG signals.”
EEG-Based Brain-Computer Interfaces (BCIs): A Survey of Recent Studies on Signal Sensing Technologies and Computational Intelligence Approaches and Their Applications (2020, background): “MI indicates imaging executing movement of a body part rather than conducting actual body movement in BCI systems [106].”
Applications of Deep Learning and Reinforcement Learning to Biological Data (2017, methodology): “CNN was applied in the classification pipeline using – augmented common spatial pattern features which covered various frequency ranges [172]; features based on combined selective location, time, and frequency attributes which were then classified using DA [173]; and signal’s dynamic energy represen”
Subject-Independent Brain–Computer Interfaces Based on Deep Convolutional Neural Networks (2019, methodology): “To date, DL has been applied to MI-based BCIs and contributed to performance improvements [48]–[51], but they are all still limited to subject-dependent BCI.”
EEGNet: a compact convolutional neural network for EEG-based brain–computer interfaces (2016, methodology): “For example, the model proposed in [36] uses the time-frequency decomposition of channels Cz, C3 and C4 as the inputs for a motor imagery classification task.”
An Efficient LSTM Network for Emotion Recognition From Multichannel EEG Signals (2020, methodology): “related features using the input of time, frequency, and electrode location features was proposed in [40].”
Deep learning for electroencephalogram (EEG) classification tasks: a review (2019, methodology): “Tabar and Halici [43] compared the accuracies between a CNN/SAE architecture against a CNN and an SAE and found that the hybrid architecture significantly outperformed against the rest.”
EEG-Based Brain-Computer Interfaces Using Motor-Imagery: Techniques and Challenges (2019, methodology): “For example, in [55], a CNN was used for feature extraction while a SAE was used for classification, and in [56], a CNN was used to extract features that were invariant to spectral and spatial changes, whist a LSTM-RNN was used to extract temporal information.”
Current Status, Challenges, and Possible Solutions of EEG-Based Brain-Computer Interface: A Comprehensive Review (2020, methodology): “The most widespread approaches are short-time Fourier transform (STFT) (Tabar and Halici, 2017; Chaudhary et al., 2019; Ha and Jeong, 2019; Tian and Liu, 2019), continuous wavelet transform (CWT) (Borisoﬀ et al., 2004; Lee and Choi, 2019; Ieracitano et al., 2020), discrete wavelet transform (DWT)…”
Deep learning techniques for classification of electroencephalogram (EEG) motor imagery (MI) signals: a review (2021, methodology): “[98] DS1: BCI-C IV-2b [101] DS2: BCI-C II-3 [102] DS1: CV (10 folds) DS2: HO (50: 50) DS1: 77.”
Deep learning-based electroencephalography analysis: a systematic review (2019, methodology): “The most frequent model inspection techniques involved the analysis of the trained model’s weights [41, 96, 97, 120, 133, 147, 182, 189, 201, 220, 223, 230, 246].”