EpiScreen: Early Epilepsy Detection from Electronic Health Records with Large Language Models
2026-03-30 • Computation and Language
Computation and Language
AI summaryⓘ
The authors created a tool called EpiScreen that helps doctors detect epilepsy early by analyzing regular clinical notes from patients' electronic health records. This tool uses advanced language models trained on medical data to tell epilepsy apart from similar conditions that look like seizures but need different treatments. EpiScreen showed strong accuracy in tests and helped doctors make better diagnoses when used together. Their work suggests that EpiScreen could speed up diagnosis and reduce unnecessary treatments, especially where expensive tests aren't easy to access.
epilepsypsychogenic non-epileptic seizuresvideo-electroencephalographyelectronic health recordslarge language modelsMIMIC-IV datasetdiagnostic delayAUCneurologistclinical notes
Authors
Shuang Zhou, Kai Yu, Zaifu Zhan, Huixue Zhou, Min Zeng, Feng Xie, Zhiyi Sha, Rui Zhang
Abstract
Epilepsy and psychogenic non-epileptic seizures often present with similar seizure-like manifestations but require fundamentally different management strategies. Misdiagnosis is common and can lead to prolonged diagnostic delays, unnecessary treatments, and substantial patient morbidity. Although prolonged video-electroencephalography is the diagnostic gold standard, its high cost and limited accessibility hinder timely diagnosis. Here, we developed a low-cost, effective approach, EpiScreen, for early epilepsy detection by utilizing routinely collected clinical notes from electronic health records. Through fine-tuning large language models on labeled notes, EpiScreen achieved an AUC of up to 0.875 on the MIMIC-IV dataset and 0.980 on a private cohort of the University of Minnesota. In a clinician-AI collaboration setting, EpiScreen-assisted neurologists outperformed unaided experts by up to 10.9%. Overall, this study demonstrates that EpiScreen supports early epilepsy detection, facilitating timely and cost-effective screening that may reduce diagnostic delays and avoid unnecessary interventions, particularly in resource-limited regions.