A deep learning approach to phase-space analysis for seizure detection

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

Many epileptic patients do not respond to medication or surgery. Recent technology has demonstrated that closed-loop responsive neurostimulation therapy is a realistic treatment for epileptic patients. However, ambulatory care of epileptic patients requires a highly accurate automated seizure detection algorithm. In this research, we implement a method for epileptic seizure detection based on nonlinear phase space analysis and deep convolutional neural networks (CNN). The underlying dynamics of scalp electroencephalography (sEEG) are extracted through time delay embedding and phase-space reconstruction. These features are used for training a CNN with a regression output to predict time until seizure. In experiments using EEG data collected in clinical environments from forty patients, our deep learning approach achieved high accuracy in predicting time until seizure onset, with a root mean squared error (RMSE) of 14.1 minutes and adjusted R-squared of .95 on out of sample testing data.

Original languageEnglish
Title of host publicationACM-BCB 2019 - Proceedings of the 10th ACM International Conference on Bioinformatics, Computational Biology and Health Informatics
PublisherAssociation for Computing Machinery, Inc
Pages190-196
Number of pages7
ISBN (Electronic)9781450366663
DOIs
StatePublished - Sep 4 2019
Event10th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics, ACM-BCB 2019 - Niagara Falls, United States
Duration: Sep 7 2019Sep 10 2019

Publication series

NameACM-BCB 2019 - Proceedings of the 10th ACM International Conference on Bioinformatics, Computational Biology and Health Informatics

Conference

Conference10th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics, ACM-BCB 2019
Country/TerritoryUnited States
CityNiagara Falls
Period09/7/1909/10/19

Keywords

  • Deep Learning
  • Epilepsy
  • Phase-Space
  • Seizure Detection

Fingerprint

Dive into the research topics of 'A deep learning approach to phase-space analysis for seizure detection'. Together they form a unique fingerprint.

Cite this