Download
Farahat_2019_J._Neural_Eng._16_066010.pdf 4,46MB
WeightNameValue
1000 Titel
  • Convolutional neural networks for decoding of covert attention focus and saliency maps for EEG feature visualization
1000 Autor/in
  1. Farahat, Amr |
  2. Reichert, Christoph |
  3. Sweeney-Reed, Catherine |
  4. Hinrichs, Hermann |
1000 Erscheinungsjahr 2019
1000 LeibnizOpen
1000 Publikationstyp
  1. Artikel |
1000 Online veröffentlicht
  • 2019-10-23
1000 Erschienen in
1000 Quellenangabe
  • 16(6):066010
1000 FRL-Sammlung
1000 Copyrightjahr
  • 2019
1000 Lizenz
1000 Verlagsversion
  • https://doi.org/10.1088/1741-2552/ab3bb4 |
1000 Publikationsstatus
1000 Begutachtungsstatus
1000 Sprache der Publikation
1000 Abstract/Summary
  • OBJECTIVE: Convolutional neural networks (CNNs) have proven successful as function approximators and have therefore been used for classification problems including electroencephalography (EEG) signal decoding for brain-computer interfaces (BCI). Artificial neural networks, however, are considered black boxes, because they usually have thousands of parameters, making interpretation of their internal processes challenging. Here we systematically evaluate the use of CNNs for EEG signal decoding and investigate a method for visualizing the CNN model decision process. APPROACH: We developed a CNN model to decode the covert focus of attention from EEG event-related potentials during object selection. We compared the CNN and the commonly used linear discriminant analysis (LDA) classifier performance, applied to datasets with different dimensionality, and analyzed transfer learning capacity. Moreover, we validated the impact of single model components by systematically altering the model. Furthermore, we investigated the use of saliency maps as a tool for visualizing the spatial and temporal features driving the model output. MAIN RESULTS: The CNN model and the LDA classifier achieved comparable accuracy on the lower-dimensional dataset, but CNN exceeded LDA performance significantly on the higher-dimensional dataset (without hypothesis-driven preprocessing), achieving an average decoding accuracy of 90.7% (chance level  =  8.3%). Parallel convolutions, tanh or ELU activation functions, and dropout regularization proved valuable for model performance, whereas the sequential convolutions, ReLU activation function, and batch normalization components reduced accuracy or yielded no significant difference. Saliency maps revealed meaningful features, displaying the typical spatial distribution and latency of the P300 component expected during this task. SIGNIFICANCE: Following systematic evaluation, we provide recommendations for when and how to use CNN models in EEG decoding. Moreover, we propose a new approach for investigating the neural correlates of a cognitive task by training CNN models on raw high-dimensional EEG data and utilizing saliency maps for relevant feature extraction.
1000 Fächerklassifikation (DDC)
1000 Liste der Beteiligten
  1. https://orcid.org/0000-0002-2141-8217|https://orcid.org/0000-0002-8649-9791|https://orcid.org/0000-0002-3684-1245|https://frl.publisso.de/adhoc/uri/SGlucmljaHMsIEhlcm1hbm4=
1000 Label
1000 Förderer
  1. Bundesministerium für Bildung und Forschung |
1000 Fördernummer
  1. 13GW0095D
1000 Förderprogramm
  1. -
1000 Dateien
1000 Förderung
  1. 1000 joinedFunding-child
    1000 Förderer Bundesministerium für Bildung und Forschung |
    1000 Förderprogramm -
    1000 Fördernummer 13GW0095D
1000 Objektart article
1000 Beschrieben durch
1000 @id frl:6417158.rdf
1000 Erstellt am 2019-10-29T11:49:38.934+0100
1000 Erstellt von 242
1000 beschreibt frl:6417158
1000 Bearbeitet von 122
1000 Zuletzt bearbeitet Wed Sep 30 17:21:52 CEST 2020
1000 Objekt bearb. Fri Feb 28 08:21:34 CET 2020
1000 Vgl. frl:6417158
1000 Oai Id
  1. oai:frl.publisso.de:frl:6417158 |
1000 Sichtbarkeit Metadaten public
1000 Sichtbarkeit Daten public
1000 Gegenstand von

View source