Learning Residual Kinematic Corrections for Continuous Neural Decoding via Reinforcement Learning

2026-07-13

AI Summary

This study addresses the limitations of non-invasive brain–computer interfaces in decoding continuous three-dimensional motor imagery, which are primarily hindered by signal variability and systematic residual errors. To overcome these challenges, the authors propose a two-stage decoding framework: an initial CNN-LSTM model generates a preliminary motion trajectory, followed by a novel application of offline reinforcement learning to correct the prediction residuals—without requiring additional neural data. Crucially, this post-processing step operates on the predicted trajectory rather than raw EEG signals. Evaluated on both 2D and virtual reality tasks, the approach achieves substantial improvements, increasing Pearson correlation coefficients by 41.5% and 21.2%, and reducing RMSE by 40.2% and 38.2%, respectively, significantly outperforming the standalone CNN-LSTM model. These results demonstrate the efficacy and innovation of offline reinforcement learning for residual correction in EEG-based decoding.

Abstract

Decoding continuous three-dimensional (3D) motor imagery (MI) using non-invasive electroencephalography (EEG)-based brain--computer interfaces (BCIs) remains challenging due to signal variability and residual decoding errors. Deep learning architectures such as convolutional neural network--long short-term memory (CNN--LSTM) models can capture spatial and temporal dynamics for continuous kinematic decoding; however, systematic residual errors persist in predicted trajectories. We propose a two-stage decoding framework that applies reinforcement learning (RL) to perform residual kinematic correction on the outputs of a CNN--LSTM decoder (CNN--LSTM--RL). The RL agent is trained offline without direct EEG input and instead operates on predicted kinematic trajectories to optimize movement accuracy relative to target trajectories. Decoding performance was quantified using Pearson correlation coefficients ($r$) and Root Mean Square Errors (RMSE) along the $x, y$, and $z$ axes. Compared to CNN--LSTM applied alone, CNN--LSTM--RL improved the mean correlation from $0.5076$ to $0.7181$ ($p = 0.0005$) in 2D and from $0.6420$ to $0.7780$ ($p = 0.0059$) in VR, with relative gains of $41.5\%$ and $21.2\%$, respectively. Correspondingly, RMSE was reduced from $0.0890$ to $0.0532$ (2D, $p < 0.0001$) and from $0.0714$ to $0.0441$ (VR, $p < 0.0001$), representing relative reductions of $40.2\%$ and $38.2\%$. These findings demonstrate that this scalable framework enhances 3D BCI MI decoding by correcting kinematic errors via offline residual RL without extra neural data, advancing neurorehabilitation, prosthetics, and virtual interaction.

Problem

• motor imagery
• EEG-based BCI
• continuous decoding
• residual errors
• kinematic decoding

Innovation

• reinforcement learning
• residual correction
• continuous neural decoding
• EEG-based BCI
• kinematic trajectory