Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

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

Mechanisms of premotor-motor cortex interactions during movement initiation.

Cell reports·2026
Same author

Efficient Gaussian process-based motor hotspot hunting with concurrent optimization of TMS coil location and orientation.

PLoS computational biology·2026
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

Efficient prospective electric field-informed localization of motor cortical targets of transcranial magnetic stimulation.

Imaging neuroscience (Cambridge, Mass.)·2025
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

Related Experiment Video

Updated: Jun 23, 2025

Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms
08:51

Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms

Published on: November 1, 2019

5.6K

Reaching the ceiling? Empirical scaling behaviour for deep EEG pathology classification.

Ann-Kathrin Kiessner1, Robin T Schirrmeister2, Joschka Boedecker3

  • 1Neuromedical AI Lab, Department of Neurosurgery, Medical Center - University of Freiburg, Faculty of Medicine, University of Freiburg, Engelbergerstr. 21, 79106, Freiburg, Germany; Neurorobotics Lab, Computer Science Department - University of Freiburg, Faculty of Engineering, University of Freiburg, Georges-Koehler-Allee 80, 79110, Freiburg, Germany; BrainLinks-BrainTools, Institute for Machine-Brain Interfacing Technology, University of Freiburg, Georges-Koehler-Allee 201, 79110, Freiburg, Germany.

Computers in Biology and Medicine
|June 15, 2024
PubMed
Summary

Deep convolutional neural networks (ConvNets) show predictable performance scaling for electroencephalogram (EEG) analysis. Understanding this scaling behavior is crucial for advancing automated EEG diagnostics.

Keywords:
AI-based clinical decision makingConvolutional Neural NetworksDeep learning in medicineEEGPathology classificationScaling laws

More Related Videos

Applications of EEG Neuroimaging Data: Event-related Potentials, Spectral Power, and Multiscale Entropy
11:15

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

Published on: June 27, 2013

33.7K
Cortical Source Analysis of High-Density EEG Recordings in Children
09:32

Cortical Source Analysis of High-Density EEG Recordings in Children

Published on: June 30, 2014

21.3K

Related Experiment Videos

Last Updated: Jun 23, 2025

Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms
08:51

Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms

Published on: November 1, 2019

5.6K
Applications of EEG Neuroimaging Data: Event-related Potentials, Spectral Power, and Multiscale Entropy
11:15

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

Published on: June 27, 2013

33.7K
Cortical Source Analysis of High-Density EEG Recordings in Children
09:32

Cortical Source Analysis of High-Density EEG Recordings in Children

Published on: June 30, 2014

21.3K

Area of Science:

  • Computational neuroscience
  • Medical imaging analysis
  • Machine learning in healthcare

Background:

  • Clinical electroencephalogram (EEG) analysis is time-consuming.
  • Deep convolutional neural networks (ConvNets) show promise for automating EEG analysis.
  • Limited understanding exists on how training data and model size affect ConvNet performance in EEG analysis.

Purpose of the Study:

  • To investigate the empirical scaling behavior of ConvNets for EEG pathology classification.
  • To analyze the impact of training sample size and model parameters on testing error.
  • To provide recommendations for optimizing deep learning models in EEG diagnostics.

Main Methods:

  • Conducted a comprehensive study on the scaling behavior of four ConvNet architectures.
  • Analyzed testing error with increasing training samples and model size (up to 1.8 million parameters).
  • Evaluated models on two large, publicly available EEG datasets (TUH Abnormal EEG Corpus and TUH Abnormal Expansion Balanced EEG Corpus).

Main Results:

  • Testing error follows a saturating power-law with both model and dataset size.
  • This scaling pattern is consistent across different ConvNet architectures and datasets.
  • Empirically observed accuracies plateau around 85%-87%, potentially limited by inter-rater agreement on clinical labels.

Conclusions:

  • The study reveals predictable scaling relationships for deep ConvNets in EEG pathology classification.
  • Findings offer valuable insights for researchers and practitioners in automated EEG diagnostics.
  • Optimizing model and dataset size based on identified scaling laws can enhance diagnostic performance.