🤖 AI Summary
Automatic EEG-based seizure detection faces significant challenges due to substantial inter-patient variability, diverse recording conditions, and the complexity of clinical environments. To address these issues, this work proposes LookAroundNet, a Transformer-based model that mimics clinicians’ interpretation strategy by incorporating contextual information from an extended temporal window surrounding the target EEG segment. The method innovatively enhances temporal context modeling and leverages multi-dataset joint training combined with model ensembling. Evaluated across multiple public and private EEG datasets, LookAroundNet demonstrates superior performance, significantly improving cross-scenario generalization while maintaining clinically feasible computational efficiency, thereby enhancing its practical deployment potential.
📝 Abstract
Automated seizure detection from electroencephalography (EEG) remains difficult due to the large variability of seizure dynamics across patients, recording conditions, and clinical settings. We introduce LookAroundNet, a transformer-based seizure detector that uses a wider temporal window of EEG data to model seizure activity. The seizure detector incorporates EEG signals before and after the segment of interest, reflecting how clinicians use surrounding context when interpreting EEG recordings. We evaluate the proposed method on multiple EEG datasets spanning diverse clinical environments, patient populations, and recording modalities, including routine clinical EEG and long-term ambulatory recordings, in order to study performance across varying data distributions. The evaluation includes publicly available datasets as well as a large proprietary collection of home EEG recordings, providing complementary views of controlled clinical data and unconstrained home-monitoring conditions. Our results show that LookAroundNet achieves strong performance across datasets, generalizes well to previously unseen recording conditions, and operates with computational costs compatible with real-world clinical deployment. The results indicate that extended temporal context, increased training data diversity, and model ensembling are key factors for improving performance. This work contributes to moving automatic seizure detection models toward clinically viable solutions.