Deep learning with convolutional neural networks for EEG decoding and visualization

Schirrmeister, Robin Tibor; Springenberg, Jost Tobias; Fiederer, Lukas Dominique Josef; Glasstetter, Martin; Eggensperger, Katharina; Tangermann, Michael; Hutter, Frank; Burgard, Wolfram; Ball, Tonio

doi:10.1002/hbm.23730

Cited by 2,108 publications

(2,387 citation statements)

References 88 publications

Supporting

Mentioning

2,128

Contrasting

Unclassified

Order By: Relevance

“…We used two convolutional network architectures, for both of which we recently showed that they decode task-related information from raw time-domain EEG with at least as good accuracies as previous state-of-the-art algorithms relying on hand-engineered features [11]. Our deep ConvNet is a fairly generic architecture (Fig.…”

Section: A Eeg Convnet Architectures and Trainingmentioning

confidence: 99%

“…2). For more details on these models, see [11]. To accommodate the longer duration of the EEG inputs as compared to our previous study, we adapted the architectures by changing the final layer filter length so the ConvNets have an input length of about 600 input samples, which correspond to 6 seconds for the 100 Hz EEG input.…”

Section: A Eeg Convnet Architectures and Trainingmentioning

confidence: 99%

“…To understand class-specific spectral characteristics in the EEG recordings, we analyzed band powers in five frequency ranges: delta (0-4 Hz), theta (4-8 Hz), alpha (8)(9)(10)(11)(12)(13)(14), low beta (14)(15)(16)(17)(18)(19)(20), high beta (20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) and low gamma .…”

Section: Visualizations Of the Spectral Differences Between Normalmentioning

confidence: 99%

“…However, the decrease is below 0.5% and thus should be interpreted cautiously. The models discovered by automated architecture optimization were markedly different from our original deep and shallow ConvNets, which were designed based on the experience in a previous study on decoding of task-related information from EEG [11]. For example, the optimized architectures used only 1.8 and 3.7 seconds of EEG data for the optimized deep and shallow ConvNet, respectively, in contrast to about 6 seconds in the original versions.…”

Section: B Deep Convnet Reached Best Accuracies Using Only 1 Minutementioning

confidence: 99%

“…These ConvNets exploit the hierarchical structure present in many natural signals. Recently, deep ConvNets trained end-to-end were, for example, able to more accurately diagnose skin cancer types from images than human dermatologists [9] and could segment retinal vessels better than human annotators [10].Deep ConvNets are nowadays also being applied to EEG analyses, such as decoding task-related information from EEG [11][12][13][14][15][16]. We have recently developed and validated the Braindecode toolbox1 for this purpose, and showed that the performance of deep ConvNets trained end-to-end is comparable to that of algorithms using hand-engineered features to decode task-related information.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Deep learning with convolutional neural networks for decoding and visualization of EEG pathology

Schirrmeister

Gemein

Eggensperger

et al. 2017

2017 IEEE Signal Processing in Medicine and Biology Symposium (SPMB)

142

View full text Add to dashboard Cite

Abstract-We apply convolutional neural networks (ConvNets) to the task of distinguishing pathological from normal EEG recordings in the Temple University Hospital EEG Abnormal Corpus. We use two basic, shallow and deep ConvNet architectures recently shown to decode task-related information from EEG at least as well as established algorithms designed for this purpose. In decoding EEG pathology, both ConvNets reached substantially better accuracies (about 6% better, ≈85% vs. ≈79%) than the only published result for this dataset, and were still better when using only 1 minute of each recording for training and only six seconds of each recording for testing. We used automated methods to optimize architectural hyperparameters and found intriguingly different ConvNet architectures, e.g., with max pooling as the only nonlinearity. Visualizations of the ConvNet decoding behavior showed that they used spectral power changes in the delta (0-4 Hz) and theta (4-8 Hz) frequency range, possibly alongside other features, consistent with expectations derived from spectral analysis of the EEG data and from the textual medical reports. Analysis of the textual medical reports also highlighted the potential for accuracy increases by integrating contextual information, such as the age of subjects. In summary, the ConvNets and visualization techniques used in this study constitute a next step towards clinically useful automated EEG diagnosis and establish a new baseline for future work on this topic.I. I Electroencephalography (EEG) is widely used in clinical practice because of its low cost and its lack of side effects due to its noninvasive nature. It is important both as a screening method as well as for hypothesis-based diagnostics, e.g., in epilepsy or stroke. One of the main limitations of using EEG for diagnostics is the required time and specialized knowledge of experts that need to be well-trained on EEG diagnostics to reach reliable results. Therefore, a machine-learning approach that aids in the diagnostic process could make EEG diagnosis more widely accessible, reduce time and effort for clinicians and potentially make diagnoses more accurate.In recent years researchers have increasingly addressed the field of computer-aided EEG diagnosis. So far, the applications were mostly limited to specific diagnoses such as Alzheimer's disease regression, neural networks, and more. This large variety of used methods indicates that the search for the best decoding approach for diverse types of EEG diagnosis is still ongoing.To overcome the lack of large datasets representative of the variety of EEG-diagnosable diseases and the heterogeneity of clinical populations, the Temple University Hospital (TUH) has published an unprecedented public dataset of clinical EEG recordings [6]. From this dataset with over 16000 clinical recordings, the TUH Abnormal EEG Corpus with about 3000 recordings has been created specifically to foster the development of methods for distinguishing pathological from normal EEG. Due to its size and rich annotatio...

show abstract

Section: A Eeg Convnet Architectures and Trainingmentioning

confidence: 99%

Section: A Eeg Convnet Architectures and Trainingmentioning

confidence: 99%

Section: Visualizations Of the Spectral Differences Between Normalmentioning

confidence: 99%

Section: B Deep Convnet Reached Best Accuracies Using Only 1 Minutementioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Deep learning with convolutional neural networks for decoding and visualization of EEG pathology

Schirrmeister

Gemein

Eggensperger

et al. 2017

2017 IEEE Signal Processing in Medicine and Biology Symposium (SPMB)

142

View full text Add to dashboard Cite

show abstract

Impact of IoT in Biomedical Applications Using Machine and Deep Learning

Rayan

Zafar

Rajab

et al. 2022

Machine Learning Algorithms for Signal and Image Processing

View full text Add to dashboard Cite

Wireless Soft Scalp Electronics and Virtual Reality System for Motor Imagery‐Based Brain–Machine Interfaces

et al. 2021

View full text Add to dashboard Cite

Motor imagery offers an excellent opportunity as a stimulus-free paradigm for brain-machine interfaces. Conventional electroencephalography (EEG) for motor imagery requires a hair cap with multiple wired electrodes and messy gels, causing motion artifacts. Here, a wireless scalp electronic system with virtual reality for real-time, continuous classification of motor imagery brain signals is introduced. This low-profile, portable system integrates imperceptible microneedle electrodes and soft wireless circuits. Virtual reality addresses subject variance in detectable EEG response to motor imagery by providing clear, consistent visuals and instant biofeedback. The wearable soft system offers advantageous contact surface area and reduced electrode impedance density, resulting in significantly enhanced EEG signals and classification accuracy. The combination with convolutional neural network-machine learning provides a real-time, continuous motor imagery-based brain-machine interface. With four human subjects, the scalp electronic system offers a high classification accuracy (93.22 ± 1.33% for four classes), allowing wireless, real-time control of a virtual reality game.

show abstract

Deep learning with convolutional neural networks for EEG decoding and visualization

Cited by 2,108 publications

References 88 publications

Deep learning with convolutional neural networks for decoding and visualization of EEG pathology

Deep learning with convolutional neural networks for decoding and visualization of EEG pathology

Impact of IoT in Biomedical Applications Using Machine and Deep Learning

Wireless Soft Scalp Electronics and Virtual Reality System for Motor Imagery‐Based Brain–Machine Interfaces

Contact Info

Product

Resources

About