Temporal Graph Convolutional Networks for Automatic Seizure Detection

TL;DR

This paper introduces the Temporal Graph Convolutional Network (TGCN), a novel deep learning model that leverages structural information in EEG data to improve automatic seizure detection, offering interpretability and competitive performance.

Contribution

The paper proposes TGCN, a new model that incorporates structural information in EEG analysis for seizure detection, with fewer parameters and enhanced interpretability.

Findings

TGCN matches state-of-the-art performance in seizure detection.

TGCN provides interpretability for clinicians to identify seizure onset and involved brain regions.

TGCN demonstrates effective use of structural information in EEG time series.

Abstract

Seizure detection from EEGs is a challenging and time consuming clinical problem that would benefit from the development of automated algorithms. EEGs can be viewed as structural time series, because they are multivariate time series where the placement of leads on a patient's scalp provides prior information about the structure of interactions. Commonly used deep learning models for time series don't offer a way to leverage structural information, but this would be desirable in a model for structural time series. To address this challenge, we propose the temporal graph convolutional network (TGCN), a model that leverages structural information and has relatively few parameters. TGCNs apply feature extraction operations that are localized and shared over both time and space, thereby providing a useful inductive bias in tasks where one expects similar features to be discriminative across the different sequences. In our experiments we focus on metrics that are most important to seizure detection, and demonstrate that TGCN matches the performance of related models that have been shown to be state of the art in other tasks. Additionally, we investigate interpretability advantages of TGCN by exploring approaches for helping clinicians determine when precisely seizures occur, and the parts of the brain that are most involved.