EEG-based Classification of Epileptic and Non-epileptic Events using Multi-array Decomposition

Abstract : In this paper, the classification of epileptic and non-epileptic events from multi-channel 16 EEG data is investigated based on temporal and spectral analysis and two different schemes for the 17 formulation of the training set. Although matrix representation which treats EEG features as 18 concatenated vectors allows capturing dependencies across EEG channels, it leads to significant 19 increase of feature vector dimensionality and lacks a means of modeling dependencies between 20 features. Thus in this paper, we compare the commonly used matrix representation in which features 21 are concatenated from all channels in order to capture the total spatiotemporal context with a tensor-22 based scheme which extracts signature features to feed the classification models. TUCKER 23 decomposition is applied to learn the essence of original, high-dimensional domain of feature space and 24 extract a multi-linear discriminative subspace. In contrast to relevant studies found in the literature, in 25 this study, the non-epileptic class consists of two types of paroxysmal episodes of loss of 26 consciousness, namely the psychogenic non-epileptic seizure (PNES) and the vasovagal syncope 27 (VVS). The classification schemes were evaluated on EEG epochs from 11 subjects in an inter-subject 28 cross-validation setting. The proposed tensor scheme achieved an accuracy of 97,7% which is better 29 2 compared to the spatiotemporal model even after trying to improve the latter by dimensionality 1 reduction through principal component analysis (PCA) and feature selection by feature ranking. 2
Liste complète des métadonnées

Cited literature [14 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01359125
Contributor : Evangelia Zacharaki <>
Submitted on : Thursday, September 1, 2016 - 7:27:48 PM
Last modification on : Thursday, February 7, 2019 - 5:29:20 PM
Document(s) archivé(s) le : Sunday, December 4, 2016 - 1:21:35 AM

File

IJMSTR_manuscript.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01359125, version 1

Citation

Evangelia Pippa, Vasileios G. Kanas, Evangelia I. Zacharaki, Vasiliki Tsirka, Michael Koutroumanidis, et al.. EEG-based Classification of Epileptic and Non-epileptic Events using Multi-array Decomposition. International Journal of Monitoring and Surveillance Technologies Research, 2017. ⟨hal-01359125⟩

Share

Metrics

Record views

377

Files downloads

458