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

Computerized EEG pattern classification by adaptive segmentation and probability density function classification.

O D Creutzfeldt, G Bodenstein, J S Barlow

    Electroencephalography and Clinical Neurophysiology
    |May 1, 1985
    PubMed
    Summary

    Computer analysis of electroencephalograms (EEGs) using adaptive segmentation and unsupervised clustering shows promise for summarizing complex brain activity. However, some limitations suggest minimal human supervision may still be necessary for accurate EEG pattern classification.

    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

    Production response to corn silage produced from normal, brown midrib, or waxy corn hybrids.

    Journal of dairy science·2012
    Same author

    The effect of age on the concentration of chloride and water in chick tissues.

    Journal of gerontology·2010
    Same author

    8-aminoquinolines as anticoccidials - Part III.

    Bioorganic & medicinal chemistry letters·1999
    Same author

    8-Aminoquinolines as anticoccidials--II.

    Bioorganic & medicinal chemistry letters·1999
    Same author

    Some aspects of alpha activity in relation to a new generalized model for the EEG.

    International journal of psychophysiology : official journal of the International Organization of Psychophysiology·1997
    Same author

    The early history of EEG data-processing at the Massachusetts Institute of Technology and the Massachusetts General Hospital.

    International journal of psychophysiology : official journal of the International Organization of Psychophysiology·1997

    Area of Science:

    • Computational Neuroscience
    • Signal Processing
    • Clinical Neurophysiology

    Background:

    • Clinical electroencephalograms (EEGs) contain complex patterns requiring accurate analysis.
    • Automated analysis of EEG data is crucial for efficient clinical interpretation.
    • Previous methods for EEG pattern classification have limitations in capturing subtle nuances.

    Purpose of the Study:

    • To analyze clinical EEGs using adaptive segmentation and unsupervised clustering.
    • To evaluate the effectiveness of computer algorithms in identifying and classifying EEG patterns.
    • To compare unsupervised clustering results with supervised hierarchical clustering and expert classification.

    Main Methods:

    • Analyzed 63 clinical EEGs using adaptive segmentation based on autocorrelation functions.

    Related Experiment Videos

  • Employed unsupervised clustering via probability density function estimation in a 2D feature space (mean frequency and power).
  • Paired EEG channels were simultaneously segmented and analyzed interactively.

    • Main Results:

    • Adaptive segmentation proved satisfactory for characterizing EEG segments.
    • Unsupervised clustering correctly identified the number of clusters in 65% of records; 35% showed over/underclustering.
    • Singular events were occasionally misclassified within clusters, and subtle electroencephalographer cues may be missed by current algorithms.

    Conclusions:

    • Computer-based EEG analysis with adaptive segmentation and unsupervised clustering offers a viable method for generating EEG summaries.
    • Current algorithms may require minimal supervision to accurately classify all EEG patterns, especially subtle ones.
    • The proposed method can provide valuable EEG summaries for clinicians, potentially reducing the need to review original records.