State-Flow Coordinated Representation for MI-EEG Decoding

2026-04-09

AI Summary

Existing motor imagery electroencephalography (MI-EEG) decoding methods typically leverage either state or flow information alone, limiting their performance. This work proposes StaFlowNet, the first framework to explicitly disentangle and synergistically fuse state and flow representations. Specifically, a dual-branch architecture is employed to separately extract global state vectors and temporal flow features, while a novel state-modulated flow module dynamically refines flow feature learning using state information. This design substantially enhances feature discriminability, enabling StaFlowNet to outperform state-of-the-art methods across three public MI-EEG datasets. Ablation studies further confirm the efficacy of the proposed state modulation mechanism.

Abstract

Motor Imagery (MI) Electroencephalography (EEG) signals contain two crucial and complementary types of information: state information, which captures the global context of the task, and flow information, which captures fine-grained temporal dynamics. However, existing deep decoding models typically focus on only one of these information streams, resulting in unstable learning and sub-optimal performance. To address this, we propose the State-Flow Coordinated Network (StaFlowNet), a novel architecture that explicitly separates and coordinates state and flow information. We first employ a dual-branch design to extract the global state vector and temporal flow features separately. Critically, a novel state-modulated flow module is proposed to dynamically refine the learning of flow information. This modulated mechanism effectively integrates global context with fine-grained dynamics, thereby significantly enhancing task discriminability and decoding performance. Experiments on three public MI-EEG datasets demonstrate that StaFlowNet significantly outperforms state-of-the-art methods. Ablation studies further confirm that the state-modulated mechanism plays a crucial role in enhancing feature discriminability and overall performance.

Problem

• MI-EEG
• state information
• flow information
• decoding performance
• temporal dynamics

Innovation

• State-Flow Coordination
• State-Modulated Flow Module
• MI-EEG Decoding
• Dual-Branch Architecture
• Temporal Dynamics