Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Safeguarding AI-driven digital health - An adaptation of the Swiss cheese model for safety.

Digital health·2026

Same author

Ethical Implications of Small Language Models (SLMs) in Healthcare Applications.

Studies in health technology and informatics·2026

Same author

Automating Safety Surveillance for Software-Based Medical Devices: Insights from FDA MAUDE Data.

Studies in health technology and informatics·2026

Same author

SurgiBot: Large Language Model-Based Patient Simulator for Surgical Informed Consent Training.

Studies in health technology and informatics·2026

Same author

Long Covid and Digital Health in Switzerland: Potentials and Limitations.

Studies in health technology and informatics·2026

Same author

Emerging Digital Interventions for ADHD: An Overview of Ongoing Clinical Trials.

Studies in health technology and informatics·2026

Search research articles

Related Experiment Video

Updated: Sep 23, 2025

A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare

A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare

Published on: July 7, 2023

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

Sugeelan Selvasingham¹, Kerstin Denecke¹

¹Bern University of Applied Sciences, Bern, Switzerland.

Studies in Health Technology and Informatics

|May 16, 2022

Summary

This summary is machine-generated.

This study compared deep learning models for classifying P300 events in children. Deep LSTM models showed the highest accuracy, but no model significantly outperformed the baseline Convolutional Neural Network (CNN).

Keywords:

Classification Convolutional neural networks EEG P300

More Related Videos

Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke

Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke

Published on: July 14, 2023

Applications of EEG Neuroimaging Data: Event-related Potentials, Spectral Power, and Multiscale Entropy

Applications of EEG Neuroimaging Data: Event-related Potentials, Spectral Power, and Multiscale Entropy

Published on: June 27, 2013

Related Experiment Videos

Last Updated: Sep 23, 2025

A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare

A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare

Published on: July 7, 2023

Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke

Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke

Published on: July 14, 2023

Applications of EEG Neuroimaging Data: Event-related Potentials, Spectral Power, and Multiscale Entropy

Applications of EEG Neuroimaging Data: Event-related Potentials, Spectral Power, and Multiscale Entropy

Published on: June 27, 2013

Area of Science:

Neuroscience
Machine Learning
Cognitive Science

Background:

P300 events, or event-related potentials, are crucial brain responses during decision-making.
Accurate classification of P300 signals is vital for understanding cognitive processes.
Deep learning offers promising avenues for analyzing complex electroencephalography (EEG) data.

Purpose of the Study:

To compare the performance of various deep learning models in classifying P300 events.
To evaluate models using a large EEG dataset from children performing a decision-making task.
To identify the most effective deep learning architecture for P300 detection in this population.

Main Methods:

Evaluated Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) variants, and Attention mechanisms.
Utilized a large, publicly available EEG dataset from school-age children.
Employed Grid Search for hyperparameter optimization and Monte Carlo Cross Validation for robust performance testing.

Main Results:

Deep LSTM achieved the highest accuracy (77.1%), closely followed by the CNN baseline (76.1%).
Statistical significance tests (5x2 paired t-test) indicated no model was significantly superior to the CNN baseline.
Extensive hyperparameter tuning resulted in 30 distinct models for comparison.

Conclusions:

While Deep LSTM showed a slight performance edge, the baseline CNN remains a competitive model for P300 classification in this context.
Further exploration of advanced architectures like Inception, ResNet, and Graph Convolutional Networks is recommended.
The findings highlight the need for continued research into optimal deep learning strategies for EEG-based cognitive event detection.