ESI-GAL: EEG source imaging-based trajectory estimation for grasp and lift task.

2024-06-17

Computers in Biology and Medicine

AI Summary

To address the challenge of real-time motor-intent decoding in non-invasive brain–computer interfaces (BCIs), this paper proposes an ESI-driven dynamic source-space trajectory inversion framework. The method integrates electroencephalographic source imaging (ESI)—specifically sLORETA and distributed source reconstruction (DSR)—with Kalman filter-based trajectory modeling, incorporating subject-specific head models and cortical constraint optimization. A novel time-varying source-domain regularization strategy is introduced to enhance stability and spatiotemporal resolution for rapid hand movements. Evaluated on 12 subjects performing continuous 3D hand trajectories (grasp-and-lift), the framework achieves a mean root-mean-square error (RMSE) of 1.8 cm—representing a 37% improvement over state-of-the-art EEG-based methods—and an end-to-end latency of <120 ms. These results satisfy the stringent real-time requirements for closed-loop control of neuroprosthetics and exoskeletons.

Technology Category

• Intelligent Robots: State Estimation
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments

Problem

• EEG signals
• 3D hand motion prediction
• Exoskeleton design

Innovation

• EEG source imaging
• Deep Learning
• rEEGNet system