1D vs 2D convolutional neural networks for scalp high frequency oscillations identification

Gaëlle Milon-Harnois (1, 2) , Nisrine Jrad (1, 2) , Daniel Schang (3, 4) , Patrick van Bogaert (5, 1) , Pierre Chauvet (1, 2)

1 LARIS - Laboratoire Angevin de Recherche en Ingénierie des Systèmes
2 UCO - Université Catholique de l'Ouest
3 LERIA - Laboratoire d'Etudes et de Recherche en Informatique d'Angers
4 ÉRIS - ESEO-ÉRIS
5 CHU Angers - Centre Hospitalier Universitaire d'Angers

Scalp High Frequency Oscillations (HFOs) are promising biomarkers of epileptogenic zones. Since HFOs visual detection is strenuous, there is a real need to develop accurate HFOs automatic detectors. In this paper, we present a comparative study of two detectors: onedimensional (1D) Convolutional Neural Networks (CNN) running on High-Density Electroencephalograms signals and two dimensional (2D) CNN on time-frequency maps of those signals. Experimental results show that 1DCNN enables easy end-to-end learning of preprocessing, feature extraction and classification modules while achieving competitive performance.

Mots clés: Epilepsy; High-Density Electroencephalogram (HD-EEG); High Frequency Oscillations; HFO; Convolutional Neural Network (CNN)

Domaines: Traitement du signal et de l'image [eess.SP]

Gaëlle Milon-Harnois, Nisrine Jrad, Daniel Schang, Patrick van Bogaert, Pierre Chauvet. 1D vs 2D convolutional neural networks for scalp high frequency oscillations identification. 2022 30th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, Oct 2022, Bruges, Belgium. pp.211. ⟨hal-03947648⟩