Classifying Numbers from EEG Data – Which Neural Network Architecture Performs Best?

Selvasingham, Sugeelan; Denecke, Kerstin (2022). Classifying Numbers from EEG Data – Which Neural Network Architecture Performs Best? In: Bürkle, Thomas; Denecke, Kerstin; Holm, Jürgen; Sariyar, Murat; Lehmann, Michael (eds.) Healthcare of the future 2022. Studies in Health Technology and Informatics: Vol. 292 (pp. 103-106). Amsterdam: IOS Press 10.3233/SHTI220333

[img]
Preview
Text
SHTI-292-SHTI220333.pdf - Published Version
Available under License Creative Commons: Attribution-Noncommercial (CC-BY-NC).

Download (761kB) | Preview

This paper presents a comparison of deep learning models for classifying P300 events, i.e., event-related potentials of the brain triggered during the human decision-making process. The evaluated models include CNN, (Bi | Deep | CNN-) LSTM, ConvLSTM, LSTM + Attention. The experiments were based on a large publicly available EEG dataset of school-age children conducting the “Guess the number”-experiment. Several hyperparameter choices were experimentally investigated resulting in 30 different models included in the comparison. Ten models with good performance on the validation data set were also automatically optimized with Grid Search. Monte Carlo Cross Validation was used to test all models on test data with 30 iterations. The best performing model was the Deep LSTM with an accuracy of 77.1% followed by the baseline (CNN) 76.1%. The significance test using a 5x2 cross validation paired t-test demonstrated that no model was significantly better than the baseline. We recommend experimenting with other architectures such as Inception, ResNet and Graph Convolutional Network.

Item Type:

Book Section (Book Chapter)

Division/Institute:

School of Engineering and Computer Science > Institute for Patient-centered Digital Health
School of Engineering and Computer Science

Name:

Selvasingham, Sugeelan;
Denecke, Kerstin0000-0001-6691-396X;
Bürkle, Thomas;
Denecke, Kerstin0000-0001-6691-396X;
Holm, Jürgen;
Sariyar, Murat and
Lehmann, Michael0000-0001-9161-5112

Subjects:

R Medicine > R Medicine (General)
T Technology > T Technology (General)

ISBN:

978-1-64368-280-8

Series:

Studies in Health Technology and Informatics

Publisher:

IOS Press

Language:

English

Submitter:

Kerstin Denecke

Date Deposited:

24 May 2022 14:45

Last Modified:

25 Oct 2023 13:56

Publisher DOI:

10.3233/SHTI220333

Uncontrolled Keywords:

Convolutional neural networks, EEG, P300, Classification

ARBOR DOI:

10.24451/arbor.16981

URI:

https://arbor.bfh.ch/id/eprint/16981

Actions (login required)

View Item View Item
Provide Feedback