Skip to main navigation Skip to search Skip to main content

FoME: A foundation model for EEG using adaptive temporal-lateral attention scaling

  • Northwestern Polytechnical University Xian
  • Chinese University of Hong Kong

Research output: Contribution to journal › Article › peer-review

1 Scopus citations

Abstract

Electroencephalography (EEG) is a vital tool to measure and record brain activity in neuroscience and clinical applications, yet its potential is constrained by signal heterogeneity, low signal-to-noise ratios, and limited labeled datasets. In this paper, we propose FoME (Foundation Model for EEG), a novel approach using adaptive temporal-lateral attention scaling to address above-mentioned challenges. FoME is pre-trained on a diverse 1.7TB dataset of scalp and intracranial EEG recordings, comprising 745M parameters trained for 1,096k steps. Our model introduces two key innovations: a time–frequency fusion embedding technique and an adaptive temporal-lateral attention scaling (ATLAS) mechanism. These components synergistically capture complex temporal and spectral EEG dynamics, enabling FoME to adapt to varying patterns across diverse data streams and facilitate robust multi-channel modeling. Evaluations across four downstream tasks demonstrate FoME’s superior performance in classification and forecasting applications, consistently achieving state-of-the-art results. To conclude, FoME establishes a new paradigm for EEG analysis, offering a versatile foundation that advances brain–computer interfaces, clinical diagnostics, and cognitive research across neuroscience and related fields. Code will be released upon publication.

Original languageEnglish
Article number102817
JournalComputerized Medical Imaging and Graphics
Volume135
DOIs
StatePublished - Oct 2026

Keywords

  • Adaptive attention scaling
  • Electroencephalography
  • Foundation model
  • Self-supervised pre-training
  • Time–frequency fusion

Fingerprint

Dive into the research topics of 'FoME: A foundation model for EEG using adaptive temporal-lateral attention scaling'. Together they form a unique fingerprint.

Cite this