The detection of the onset of focal epileptic seizures is critical for epilepsy patients to enable timely intervention and enhance their safety and quality of life. This task is profoundly challenging due to the high variability of seizures across patients, the complexity of seizure-onset patterns in the multivariate electroencephalography ({EEG}) time series, and the limited availability of annotated seizures for training machine learning models. Whereas state-of-the-art seizure detection approaches typically require annotated patient-specific data, detecting focal seizure onset for new patients unseen during model training is much more challenging and remains insufficiently explored in the literature. In this paper, we propose {PATIENT}+N \u2013 a novel approach to facilitate focal seizure-onset detection for new patients with non-annotated {EEG} data. {PATIENT}+N learns to construct feature-importance profiles for patient clusters based on their most informative features and to assign new patients to such clusters. {PATIENT}+N effectively leverages feature-importance profiles and global features to guide focal seizure-onset detection within a neural network. Our evaluation results on a real-world {EEG} dataset demonstrate that {PATIENT}+N substantially outperforms state-of-the-art baselines, achieving a 7-percentage-point increase in F1 score for the seizure-onset class on average.<\/p>\n","protected":false},"excerpt":{"rendered":"
The detection of the onset of focal epileptic seizures is critical for epilepsy patients to enable timely intervention and enhance their safety and quality of life. This task is profoundly challenging due to the high variability of seizures across patients, the complexity of seizure-onset patterns in the multivariate electroencephalography ({EEG}) time series, and the limited availability of annotated seizures for training machine learning models. Whereas state-of-the-art seizure detection approaches typically […]<\/p>\n","protected":false},"author":14,"featured_media":0,"template":"","meta":{"_acf_changed":false,"footnotes":""},"publication-type":[32],"publication-year":[1679],"class_list":["post-43768","publication","type-publication","status-publish","hentry","publication-type-inproceedings","publication-year-1679"],"acf":[],"publishpress_future_workflow_manual_trigger":{"enabledWorkflows":[]},"_links":{"self":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication\/43768","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication"}],"about":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/types\/publication"}],"author":[{"embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/users\/14"}],"version-history":[{"count":0,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication\/43768\/revisions"}],"wp:attachment":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/media?parent=43768"}],"wp:term":[{"taxonomy":"publication-type","embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication-type?post=43768"},{"taxonomy":"publication-year","embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication-year?post=43768"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}