From Epilepsy Seizures Classification to Detection: A Deep Learning-based Approach for Raw EEG Signals

TL;DR

This paper presents a deep learning pipeline for detecting epileptic seizures from raw EEG signals, incorporating novel preprocessing, post-processing, and evaluation techniques, demonstrating high accuracy and cross-species generalization.

Contribution

The study introduces a comprehensive seizure detection pipeline with innovative preprocessing, post-processing, and evaluation methods, and demonstrates cross-species generalization of the model.

Findings

Achieved 93% F1-score on human EEGs using animal-trained model.

Developed a new segmentation and reassembly approach for raw EEG signals.

Showed the difference in performance between seizure classification and detection.

Abstract

Epilepsy represents the most prevalent neurological disease in the world. One-third of people suffering from mesial temporal lobe epilepsy (MTLE) exhibit drug resistance, urging the need to develop new treatments. A key part in anti-seizure medication (ASM) development is the capability of detecting and quantifying epileptic seizures occurring in electroencephalogram (EEG) signals, which is crucial for treatment efficacy evaluation. In this study, we introduced a seizure detection pipeline based on deep learning models applied to raw EEG signals. This pipeline integrates: a new pre-processing technique which segments continuous raw EEG signals without prior distinction between seizure and seizure-free activities; a post-processing algorithm developed to reassemble EEG segments and allow the identification of seizures start/end; and finally, a new evaluation procedure based on a strict seizure events comparison between predicted and real labels. Models training have been performed using a data splitting strategy which addresses the potential for data leakage. We demonstrated the fundamental differences between a seizure classification and a seizure detection task and showed the differences in performance between the two tasks. Finally, we demonstrated the generalization capabilities across species of our best architecture, combining a Convolutional Neural Network and a Transformer encoder. The model was trained on animal EEGs and tested on human EEGs with a F1-score of 93% on a balanced Bonn dataset.