Evaluation of novel algorithm embedded in a wearable sEMG device for seizure detection

Isa Conradsen, Sandor Beniczky, Peter Wolf, Poul Jennum, Helge B.D. Sorensen

    Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

    Abstract

    We implemented a modified version of a previously published algorithm for detection of generalized tonic-clonic seizures into a prototype wireless surface electromyography (sEMG) recording device. The method was modified to require minimum computational load, and two parameters were trained on prior sEMG data recorded with the device. Along with the normal sEMG recording, the device is able to set an alarm whenever the implemented algorithm detects a seizure. These alarms are annotated in the data file along with the signal. The device was tested at the Epilepsy Monitoring Unit (EMU) at the Danish Epilepsy Center. Five patients were included in the study and two of them had generalized tonic-clonic seizures. All patients were monitored for 2–5 days. A double-blind study was made on the five patients. The overall result showed that the device detected four of seven seizures and had a false detection rate of 0.003/h or one in twelve days.
    Original languageEnglish
    Title of host publicationIEEE Engineering in medicine and biology society conference proceedings
    PublisherIEEE
    Publication date2012
    Pages2048 - 2051
    ISBN (Print)9781424441198
    DOIs
    Publication statusPublished - 2012
    Event34th Annual International Conference of the IEEE EMBS - San Diego, California, United States
    Duration: 28 Aug 20121 Sept 2012

    Conference

    Conference34th Annual International Conference of the IEEE EMBS
    Country/TerritoryUnited States
    CitySan Diego, California
    Period28/08/201201/09/2012

    Fingerprint

    Dive into the research topics of 'Evaluation of novel algorithm embedded in a wearable sEMG device for seizure detection'. Together they form a unique fingerprint.

    Cite this