CLEF: EEG Foundation Model for Learning Clinical Semantics

TL;DR

CLEF is a novel long-context EEG foundation model that integrates clinical data and outperforms previous models on a comprehensive 234-task benchmark, advancing clinical EEG interpretation.

Contribution

Introduces CLEF, a clinically grounded, session-scale EEG foundation model using 3D spectrogram tokens and contrastive learning with clinical data, surpassing prior models.

Findings

CLEF outperforms prior models on 229 of 234 tasks.

Reconstruction pretraining surpasses previous EEG models.

Alignment with reports and EHR improves performance and transferability.

Abstract

Clinical EEG interpretation requires reasoning over full EEG sessions and integrating signal patterns with clinical context. Existing EEG foundation models are largely designed for short-window decoding and do not incorporate clinical context. We introduce CLEF, a clinically grounded long-context EEG foundation model. CLEF represents EEG sessions as 3D multitaper spectrogram tokens, enabling tractable Transformer modeling at session scale, and aligns embeddings with neurologist reports and structured EHR data through contrastive objectives. We evaluate CLEF on a new 234-task benchmark spanning disease phenotypes, medication exposures, and EEG findings, with more than 260k EEG sessions from over 108k patients. CLEF outperforms prior EEG foundation models on 229 of 234 tasks, improving mean AUROC from 0.65 to 0.74. Reconstruction-only pretraining surpasses prior EEG foundation models, while report and EHR alignment yields further gains. Held-out concept and external-cohort experiments suggest that these representations transfer beyond observed alignment targets. These results support session-scale, clinically grounded representation learning as a promising foundation-model paradigm for clinical EEG.