Related Experiment Video
Updated: Jul 14, 2025
10:51
An Experimental Platform to Study the Closed-loop Performance of Brain-machine Interfaces
Published on: March 10, 2011
13.8K
Audio-induced medial prefrontal cortical dynamics enhances coadaptive learning in brain-machine interfaces
Jieyuan Tan1, Xiang Zhang1, Shenghui Wu1
1Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology, Hong Kong Special Administrative Region of China, People's Republic of China.
Journal of Neural Engineering
|October 9, 2023
Summary
This study introduces a coadaptive brain-machine interface (BMI) using audio feedback to improve both subject learning and decoder performance. The system enhances learning efficiency for paralyzed individuals by leveraging neural activity for faster adaptation.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Machine Learning
Background:
- Coadaptive brain-machine interfaces (BMIs) offer solutions for paralyzed individuals by enabling mutual adaptation between users and devices.
- Existing research focuses on improving sensory feedback for user learning or adaptive algorithms for decoder stability.
Purpose of the Study:
- To develop an efficient coadaptive BMI framework that enhances subject learning via sensory feedback and improves decoder adaptability by utilizing feedback-induced evaluation information.
- To investigate the impact of dynamic audio feedback on learning performance in a behavioral task.
Main Methods:
- Dynamic audio feedback was designed based on subject performance during a new task acquisition.
- A coadaptive framework incorporated medial prefrontal cortex (mPFC) activity into the closed-loop system.
- Neural dynamics of audio-induced mPFC activity were analyzed and translated into reward expectation for decoder learning.
Main Results:
- Audio feedback improved behavioral performance, reducing the time to reach 80% proficiency by 26.4%.
- Significant neural responses in the mPFC were elicited by audio feedback, correlating with behavioral improvements.
- Decoders utilizing audio-induced reward expectation demonstrated 33.8% faster task learning on average.
Conclusions:
- Designed audio feedback effectively enhances subject learning in coadaptive BMIs.
- Utilizing mPFC activity induced by audio feedback improves decoder learning efficiency for new tasks.
- The proposed coadaptive framework advances BMIs toward greater adaptability and learning capability.
