Self-Supervised Consistency Enhanced Disentangled Learning for Neural Decoding Generalization in Brain-Machine Interface

2026-07-27

AI Summary

This study addresses the significant degradation in decoding performance of invasive brain–computer interfaces (BCIs) over time due to neural signal drift, which hinders long-term stable control. To overcome this challenge, the authors propose the SSCDL framework, which first incorporates a self-supervised consistency constraint under simulated perturbations, leveraging a teacher–student architecture to learn neural representations robust to drift. Furthermore, drawing upon neural preference theory, the framework disentangles movement parameters into speed, direction, and magnitude, enabling multi-dimensional complementary generalization. Evaluated in cross-day experiments, the method substantially improves both decoding accuracy and system stability. Notably, this work presents the first implementation of neural-preference-guided disentangled learning, offering an effective solution for achieving long-term, high-precision human–machine interaction.

Abstract

Brain-Machine Interfaces (BMIs) provide a direct communication pathway between the brain and external devices, enabling humans to control assistive and robotic technologies, with potential applications in rehabilitation, human motor augmentation, and human-centered robotics. However, due to neural drift, the performance of BMIs decreases over time, posing challenges for long-term viability, particularly for invasive BMIs (iBMIs). Existing solutions suffer from two main drawbacks: (i) difficulty in learning robust neural representations, and (ii) neglecting that neural drift varies across motor parameters (e.g., velocity, direction, and speed). To overcome these limitations, we propose Self-Supervised Consistency enhanced Disentangled Learning (SSCDL), a neural decoding generalization framework built on two key innovations. We first design a backbone model named Consistency enhanced Neural Decoder (CND), using a novel teacher-student consistency constraint with simulated neural signal perturbations to learn robust representations invariant to neural drift. Then, we employ three dedicated CNDs under the Complementary-Disentangled Generalization (CDG) mechanism, which disentangles motor signals into velocity, direction, and speed with inspiration from neural preference theory. This disentangled learning enables SSCDL to capture invariant neural representations from diverse neural preference perspectives, significantly enhancing cross-day generalization. Extensive experimental results show that SSCDL delivers state-of-the-art decoding performance, exhibiting high robustness and cross-day stability. These capabilities underscore its strong potential for long-term interaction in human-centric robotic and fine-grained assistive applications.

Problem

• neural drift
• brain-machine interface
• decoding generalization
• motor parameters
• long-term stability

Innovation

• Self-Supervised Learning
• Disentangled Representation
• Neural Decoding
• Neural Drift
• Brain-Machine Interface