Time-Varying EEG Correlations Improve Automated Neonatal Seizure Detection.

Journal: International journal of neural systems
Published Date: Jun 24, 2018

Abstract

The aim of this study was to develop methods for detecting the nonstationary periodic characteristics of neonatal electroencephalographic (EEG) seizures by adapting estimates of the correlation both in the time (spike correlation; SC) and time-frequency domain (time-frequency correlation; TFC). These measures were incorporated into a seizure detection algorithm (SDA) based on a support vector machine to detect periods of seizure and nonseizure. The performance of these nonstationary correlation measures was evaluated using EEG recordings from 79 term neonates annotated by three human experts. The proposed measures were highly discriminative for seizure detection (median : 0.933 IQR: 0.821-0.975, median : 0.883 IQR: 0.707-0.931). The resultant SDA applied to multi-channel recordings had a median AUC of 0.988 (IQR: 0.931-0.998) when compared to consensus annotations, outperformed two state-of-the-art SDAs and was noninferior to the human expert for 73/79 of neonates.

Authors

  • Karoliina T Tapani
    1 BABA Center, Children's Hospital, HUS Medical Imaging Center, Clinical Neurophysiology, University of Helsinki, Helsinki University Hospital and University of Helsinki, Finland.
  • Sampsa Vanhatalo
    1 BABA Center, Children's Hospital, HUS Medical Imaging Center, Clinical Neurophysiology, University of Helsinki, Helsinki University Hospital and University of Helsinki, Finland.
  • Nathan J Stevenson
    3 Clinicum, Faculty of Medicine, University of Helsinki, Helsinki, Finland.

Keywords

External Resources

Popular Topics
Recent Journals