Enhancing brain decoding using attention augmented deep neural networks

I Alaoui Abdellaoui; J Garcia Fernández; C Şahinli; Siamak Mehrkanoon

doi:10.14428/esann/2021.ES2021-67

Enhancing brain decoding using attention augmented deep neural networks

I Alaoui Abdellaoui, J Garcia Fernández, C Şahinli, Siamak Mehrkanoon

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

Abstract

Neuroimaging techniques have shown to be valuable when studying brain activity. This paper uses Magnetoencephalography (MEG) data, provided by the Human Connectome Project (HCP), and different deep learning models to perform brain decoding. Specifically, we investigate to which extent one can infer the task performed by a subject based on its MEG data. In order to capture the most relevant features of the signals, self and global attention are incorporated into our models. The obtained results show that the inclusion of attention improves the performance and generalization of the models across subjects.

Original language	English
Title of host publication	European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN)
Publisher	i6doc
Pages	183-188
Number of pages	6
DOIs	https://doi.org/10.14428/esann/2021.ES2021-67
Publication status	Published - 2021
Event	29th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning - Online, Belgium Duration: 6 Oct 2021 → 8 Oct 2021 https://www.esann.org/

Symposium

Symposium	29th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
Abbreviated title	ESANN 2021
Country/Territory	Belgium
Period	6/10/21 → 8/10/21
Internet address	https://www.esann.org/

Access to Document

10.14428/esann/2021.ES2021-67

https://www.esann.org/proceedings/2021

Cite this

@inproceedings{64a44bf0f6f84e89a636333f89a78eeb,

title = "Enhancing brain decoding using attention augmented deep neural networks",

abstract = "Neuroimaging techniques have shown to be valuable when studying brain activity. This paper uses Magnetoencephalography (MEG) data, provided by the Human Connectome Project (HCP), and different deep learning models to perform brain decoding. Specifically, we investigate to which extent one can infer the task performed by a subject based on its MEG data. In order to capture the most relevant features of the signals, self and global attention are incorporated into our models. The obtained results show that the inclusion of attention improves the performance and generalization of the models across subjects.",

author = "Abdellaoui, {I Alaoui} and Fern{\'a}ndez, {J Garcia} and C {\c S}ahinli and Siamak Mehrkanoon",

note = "Publisher Copyright: {\textcopyright} 2021 ESANN Intelligence and Machine Learning. All rights reserved.; 29th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2021 ; Conference date: 06-10-2021 Through 08-10-2021",

year = "2021",

doi = "10.14428/esann/2021.ES2021-67",

language = "English",

pages = "183--188",

booktitle = "European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN)",

publisher = "i6doc",

address = "Belgium",

url = "https://www.esann.org/",

}

Abdellaoui, IA, Fernández, JG, Şahinli, C & Mehrkanoon, S 2021, Enhancing brain decoding using attention augmented deep neural networks. in European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN). i6doc, pp. 183-188, 29th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, Belgium, 6/10/21. https://doi.org/10.14428/esann/2021.ES2021-67

Enhancing brain decoding using attention augmented deep neural networks. / Abdellaoui, I Alaoui; Fernández, J Garcia; Şahinli, C et al.
European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN). i6doc, 2021. p. 183-188.

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

TY - GEN

T1 - Enhancing brain decoding using attention augmented deep neural networks

AU - Abdellaoui, I Alaoui

AU - Fernández, J Garcia

AU - Şahinli, C

AU - Mehrkanoon, Siamak

PY - 2021

Y1 - 2021

N2 - Neuroimaging techniques have shown to be valuable when studying brain activity. This paper uses Magnetoencephalography (MEG) data, provided by the Human Connectome Project (HCP), and different deep learning models to perform brain decoding. Specifically, we investigate to which extent one can infer the task performed by a subject based on its MEG data. In order to capture the most relevant features of the signals, self and global attention are incorporated into our models. The obtained results show that the inclusion of attention improves the performance and generalization of the models across subjects.

AB - Neuroimaging techniques have shown to be valuable when studying brain activity. This paper uses Magnetoencephalography (MEG) data, provided by the Human Connectome Project (HCP), and different deep learning models to perform brain decoding. Specifically, we investigate to which extent one can infer the task performed by a subject based on its MEG data. In order to capture the most relevant features of the signals, self and global attention are incorporated into our models. The obtained results show that the inclusion of attention improves the performance and generalization of the models across subjects.

U2 - 10.14428/esann/2021.ES2021-67

DO - 10.14428/esann/2021.ES2021-67

M3 - Conference article in proceeding

SP - 183

EP - 188

BT - European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN)

PB - i6doc

T2 - 29th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning

Y2 - 6 October 2021 through 8 October 2021

ER -