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 Experiment Videos

Enhancing the detection of seizures with a clustering algorithm

A Klatchko1, G Raviv, W R Webber

  • 1Bio-logic Systems Corp., Mundelein, IL 60060, USA. asher@blsc.com

Electroencephalography and Clinical Neurophysiology
|July 29, 1998
PubMed
Summary

This study introduces a novel clustering algorithm to improve electroencephalogram (EEG) seizure detection. The method enhances seizure onset detection and significantly reduces false positives by analyzing detector results across time and channels.

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

GALACTIC COSMIC RAYS IN THE LOCAL INTERSTELLAR MEDIUM: <i>VOYAGER 1</i> OBSERVATIONS AND MODEL RESULTS.

The Astrophysical journal·2021
Same author

IVF outcome in azoospermic cancer survivors.

European journal of obstetrics, gynecology, and reproductive biology·2017
Same author

Correlation between premature ejaculation and female vaginal penetration difficulties.

International journal of impotence research·2015
Same author

Selective interference with the representation of size in the human by direct cortical electrical stimulation.

Journal of cognitive neuroscience·2013
Same author

Localizing arithmetic processes in the brain: evidence from a transient deficit during cortical stimulation.

Journal of cognitive neuroscience·2013
Same author

Voyager 1 observes low-energy galactic cosmic rays in a region depleted of heliospheric ions.

Science (New York, N.Y.)·2013

Area of Science:

  • Neuroscience
  • Medical Informatics
  • Signal Processing

Background:

  • Automated electroencephalogram (EEG) seizure detection algorithms often analyze limited time epochs and single channels.
  • This approach can lead to missed detections and a high false detection rate due to seizures spanning multiple electrodes and time periods.
  • Existing methods struggle with weak correspondence between actual seizure patterns and algorithm-generated event counterparts.

Purpose of the Study:

  • To develop an improved algorithm for automated EEG seizure detection.
  • To address the limitations of current algorithms, specifically false detections and missed seizure onsets.
  • To enhance the accuracy and reliability of seizure detection systems.

Main Methods:

  • Representing detector weights/probabilities as a directed graph (digraph) where seizures correspond to connected components.

Related Experiment Videos

  • Introducing a clustering algorithm that correlates detector results across both time and channels.
  • Extending detection capabilities to an unlimited number of electrodes over indefinite time.

    • Main Results:

    • The proposed algorithm effectively extends detection to multiple electrodes and time durations.
    • Demonstrated a strong correlation between detected seizure events and the boundaries of identified clusters.
    • Achieved over a 50% reduction in the false detection rate compared to generic detectors.

    Conclusions:

    • The developed clustering algorithm significantly enhances the detection of seizure onset.
    • The method effectively lowers the rate of false detections in EEG seizure analysis.
    • The algorithm is computationally efficient (linear time complexity, O(m)) and suitable for real-time implementation.