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

Biomarkers for Alzheimer's disease and mild cognitive impairment: Recent advances in task-based EEG.

Journal of Alzheimer's disease : JAD·2026

Same author

New avenues for understanding what deep networks learn from EEG.

Frontiers in robotics and AI·2025

Same author

Cardiac magnetic resonance imaging in the German National Cohort (NAKO): Automated segmentation of short-axis cine images and post-processing quality control.

Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance·2025

Same author

EEG-CLIP: learning EEG representations from natural language descriptions.

Frontiers in robotics and AI·2025

Same author

Brain age revisited: Investigating the state vs. trait hypotheses of EEG-derived brain-age dynamics with deep learning.

Imaging neuroscience (Cambridge, Mass.)·2025

Same author

Corticomuscular coherence during upright standing in unilateral transfemoral amputees.

Brain communications·2025

Search research articles

Related Experiment Video

Updated: Sep 11, 2025

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

Published on: July 26, 2013

Deep Riemannian Networks for end-to-end EEG decoding.

Daniel Wilson^1,2, Robin T Schirrmeister^1,2,3, Lukas A W Gemein¹

¹Neuromedical A.I. Lab, Department of Neurosurgery, Medical Center-University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany.

Imaging Neuroscience (Cambridge, Mass.)

|August 13, 2025

Summary

This summary is machine-generated.

Deep Riemannian Networks (DRNs) show promise for electroencephalography (EEG) decoding. Our study introduces an end-to-end DRN that outperforms traditional methods, using physiologically plausible frequency regions for improved EEG analysis.

Keywords:

EEG Riemannian and Deep Learning filterbank

More Related Videos

Neuroimaging-Guided TMS–EEG for Real-Time Cortical Network Mapping

Neuroimaging-Guided TMS–EEG for Real-Time Cortical Network Mapping

Published on: June 13, 2025

EEG Mu Rhythm in Typical and Atypical Development

EEG Mu Rhythm in Typical and Atypical Development

Published on: April 9, 2014

Related Experiment Videos

Last Updated: Sep 11, 2025

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

Published on: July 26, 2013

Neuroimaging-Guided TMS–EEG for Real-Time Cortical Network Mapping

Neuroimaging-Guided TMS–EEG for Real-Time Cortical Network Mapping

Published on: June 13, 2025

EEG Mu Rhythm in Typical and Atypical Development

EEG Mu Rhythm in Typical and Atypical Development

Published on: April 9, 2014

Area of Science:

Neuroscience
Machine Learning
Signal Processing

Background:

State-of-the-art electroencephalography (EEG) decoding relies on Deep Learning (DL) or Riemannian Geometry-based Decoders (RBDs).
Deep Riemannian Networks (DRNs) integrate DL and RBDs, but their architectural design and data transformation require further investigation for widespread EEG application.

Purpose of the Study:

To explore the impact of hyperparameters on Deep Riemannian Network (DRN) performance in EEG decoding.
To analyze data transformations within DRNs and their correlation with traditional EEG decoding methods.
To propose and evaluate an end-to-end DRN for high-performance EEG decoding.

Main Methods:

Analysis of DRNs with a wide range of hyperparameters across five public EEG datasets.
Comparison of the proposed end-to-end EEG SPDNet (EE(G)-SPDNet) with state-of-the-art Convolutional Neural Networks (ConvNets).
Investigation of filter learning and channel-specific filtering approaches within the DRN architecture.

Main Results:

The proposed EE(G)-SPDNet, a wide, end-to-end DRN, demonstrated superior performance compared to ConvNets.
The end-to-end DRN learned complex filters beyond traditional bandpass filters (alpha, beta, gamma) and utilized physiologically plausible frequency regions.
Performance gains were observed with channel-specific filtering, though architectural analysis indicated potential for improved utilization of Riemannian information.

Conclusions:

End-to-end DRNs, like EE(G)-SPDNet, can effectively infer task-related information from raw EEG without manual filterbanks.
The study provides foundational insights into designing and training DRNs for high-performance EEG decoding.
EE(G)-SPDNet highlights the potential of DRNs to advance EEG analysis and applications.