LEAD: Large Foundation Model for EEG-Based Alzheimer's Disease Detection

2025-02-02

AI Summary

To address the dual challenges of scarce large-scale labeled data and substantial inter-subject variability in Alzheimer’s disease (AD) detection from electroencephalography (EEG), this work introduces the largest publicly available EEG-AD dataset to date (813 subjects) and proposes the first EEG-oriented large foundation model for AD detection. Methodologically, we establish the first dedicated large-model architecture for EEG; design a dual-granularity self-supervised contrastive pretraining framework—operating at both sample-level and subject-level; incorporate a spatiotemporal joint embedding encoder to jointly model temporal dynamics and channel-wise topological structure; and employ cross-dataset channel alignment fine-tuning with subject-independent cross-validation. The resulting end-to-end AD classifier achieves F1-score improvements of 9.86% (sample-level) and 9.31% (subject-level) over state-of-the-art methods, demonstrating the efficacy of contrastive pretraining and channel alignment in mitigating inter-subject heterogeneity.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Large Vision Models
• Cognitive Modeling & Cognitive Systems: Other Foundations of Cognitive Modeling & Systems

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Practical large-scale studies of user experience

Abstract

Electroencephalogram (EEG) provides a non-invasive, highly accessible, and cost-effective solution for Alzheimer's Disease (AD) detection. However, existing methods, whether based on manual feature extraction or deep learning, face two major challenges: the lack of large-scale datasets for robust feature learning and evaluation, and poor detection performance due to inter-subject variations. To address these challenges, we curate an EEG-AD corpus containing 813 subjects, which forms the world's largest EEG-AD dataset to the best of our knowledge. Using this unique dataset, we propose LEAD, the first large foundation model for EEG-based AD detection. Our method encompasses an entire pipeline, from data selection and preprocessing to self-supervised contrastive pretraining, fine-tuning, and key setups such as subject-independent evaluation and majority voting for subject-level detection. We pre-train the model on 11 EEG datasets and unified fine-tune it on 5 AD datasets. Our self-supervised pre-training design includes sample-level and subject-level contrasting to extract useful general EEG features. Fine-tuning is performed on 5 channel-aligned datasets together. The backbone encoder incorporates temporal and channel embeddings to capture features across both temporal and spatial dimensions. Our method demonstrates outstanding AD detection performance, achieving up to a 9.86% increase in F1 score at the sample-level and up to a 9.31% at the subject-level compared to state-of-the-art methods. The results of our model strongly confirm the effectiveness of contrastive pre-training and channel-aligned unified fine-tuning for addressing inter-subject variation. The source code is at https://github.com/DL4mHealth/LEAD.

Problem

• Detects Alzheimer's Disease using EEG data
• Addresses lack of large EEG datasets for AD
• Improves detection accuracy across different subjects

Innovation

• Large EEG-AD dataset with 813 subjects
• Self-supervised contrastive pre-training design
• Channel-aligned unified fine-tuning for AD detection