MIRepNet: A Pipeline and Foundation Model for EEG-Based Motor Imagery Classification

2025-07-27

AI Summary

Existing EEG foundation models overlook the neurophysiological specificity of motor imagery (MI) paradigms, resulting in limited generalizability. To address this, we propose MIRepNet—the first foundation model explicitly designed for MI. It innovatively integrates neurophysiology-informed channel-template mapping with an adaptive preprocessing pipeline and employs a hybrid self-supervised (masked token reconstruction) and supervised (MI classification) pretraining strategy. MIRepNet supports arbitrary electrode configurations, enabling efficient cross-device and cross-dataset adaptation. Evaluated on five public MI datasets, it achieves state-of-the-art performance, significantly outperforming both general-purpose and existing task-specific models. Moreover, it enables rapid few-shot fine-tuning with fewer than 30 trials per class, substantially enhancing downstream generalization and clinical deployability.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Computer Vision: Multi-modal Vision

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Brain-computer interfaces (BCIs) enable direct communication between the brain and external devices. Recent EEG foundation models aim to learn generalized representations across diverse BCI paradigms. However, these approaches overlook fundamental paradigm-specific neurophysiological distinctions, limiting their generalization ability. Importantly, in practical BCI deployments, the specific paradigm such as motor imagery (MI) for stroke rehabilitation or assistive robotics, is generally determined prior to data acquisition. This paper proposes MIRepNet, the first EEG foundation model tailored for the MI paradigm. MIRepNet comprises a high-quality EEG preprocessing pipeline incorporating a neurophysiologically-informed channel template, adaptable to EEG headsets with arbitrary electrode configurations. Furthermore, we introduce a hybrid pretraining strategy that combines self-supervised masked token reconstruction and supervised MI classification, facilitating rapid adaptation and accurate decoding on novel downstream MI tasks with fewer than 30 trials per class. Extensive evaluations across five public MI datasets demonstrated that MIRepNet consistently achieved state-of-the-art performance, significantly outperforming both specialized and generalized EEG models. Our code will be available on GitHubfootnote{https://github.com/staraink/MIRepNet}.

Problem

• Develops EEG foundation model for motor imagery classification
• Addresses limitations in generalization of current EEG models
• Enhances accuracy with hybrid pretraining and adaptable pipeline

Innovation

• EEG foundation model for motor imagery
• Neurophysiologically-informed preprocessing pipeline
• Hybrid self-supervised and supervised pretraining