/ EN
2026-07-14 00:00 China Papers Foundations & Methods Translated from EN

Federated Graph Framework Fuses EEG and EMG to Decode Motor Intent

Summary A team writing in Computing and Informatics has proposed FSDFGL, a federated graph learning framework that folds EMG into EEG when the graph is built and then shares only structural information between clients, avoiding the accuracy loss that comes with exchanging heterogeneous features. The design targets three problems in hybrid BCI work at once: the low spatial resolution of EEG, datasets that are small and privacy-sensitive, and the uneven data distributions federated learning has to cope with. Experiments on the TAN dataset show an advantage in identifying complex motor intentions.
Why it matters Sharing graph structure instead of features is the move worth noting: it would let hospitals pool hybrid EEG-EMG models without moving patient signals, which is the practical obstacle to assembling BCI datasets large enough to matter.

BCIwiki (bciwiki.com) — A new study published in Computing and Informatics on July 14, 2026, proposes a federated graph learning framework named FSDFGL for motor intention recognition in hybrid brain-computer interfaces. The research was conducted by scientists from Shanghai Sunshine Rehabilitation Center and Shanghai Ocean University.

The study highlights that deep learning-based EEG decoding faces multiple challenges: EEG has low spatial resolution, making it difficult to accurately identify specific brain regions; deep learning requires large amounts of data, while EEG datasets are typically small; and merging small datasets raises privacy concerns because EEG signals contain a wealth of personal information. Federated learning (FL) offers a potential solution for limited data and privacy, but it faces the challenge of non-independent and identically distributed (non-IID) data. To address these issues, the team proposed incorporating EMG data into EEG data through SPMI during graph construction, leveraging the distinct features of muscle activity provided by EMG signals. In the local graph neural network (GNN) process, structural information is decoupled from node features, and a structure encoder captures and shares this information globally during FL. The framework was evaluated on the TAN dataset, and experimental results demonstrate its superiority in identifying complex motor intentions.

By sharing structural information rather than heterogeneous features, the framework avoids degradation of local model performance, potentially offering a new approach to data privacy and model generalization in hybrid BCIs.

Compiled by BCIwiki from public sources

Sources · 1
doi.org 2026-07-14

Shanghai Sunshine Rehabilitation Center timeline

2026-06 Multi-View Contrastive Learning Improves Cross-Subject ERP Classification 2026-04 Shanghai Yangzhi Rehabilitation Hospital Registers Trial of rTMS-Enhanced Motor Imagery BCI for Post-Stroke Upper Limb Dysfunction All entries →
Read original ↗
Suggest a correction Revisions · none / EN
© 2026 BCIwiki.com Digest Topics Tips Subscribe Revisions About