Related Experiment Video
Updated: Apr 16, 2026
10:14
Author Spotlight: Enhancing Neurorehabilitation Through EEG, Motor Imagery, and Virtual Reality
Published on: May 10, 2024
2.1K
Classification of hemodynamic responses associated with force and speed imagery for a brain-computer interface
Xuxian Yin1, Baolei Xu, Changhao Jiang
1State Key Laboratory of Robotics, Shenyang Institute of Automation (SIA), Chinese Academy of Sciences (CAS), Shenyang, 110016, Peoples Republic of China.
Journal of Medical Systems
|March 4, 2015
Summary
Functional near-infrared spectroscopy (fNIRS) brain-computer interface (BCI) systems show promise for decoding hand clenching tasks. This study achieved 76.7% accuracy, offering a new paradigm for BCI development.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Functional near-infrared spectroscopy (fNIRS) is an emerging optical technique for assessing task-associated brain activity.
- Brain-computer interfaces (BCIs) offer novel ways to interact with technology using neural signals.
Purpose of the Study:
- To develop and evaluate a 2-class fNIRS-BCI system for decoding hand clenching imageries.
- To explore multi-class fNIRS-BCI performance based on varying force and speed levels.
Main Methods:
- Six participants performed hand clenching imageries involving force and speed variations.
- Joint Mutual Information (JMI) was used for optimal feature extraction of hemodynamic responses.
- An Extreme Learning Machine (ELM) classifier was employed for its speed and parameter insensitivity.
Main Results:
- A 2-class fNIRS-BCI system achieved an average accuracy of 76.7% when force and speed tasks were broadly categorized.
- Moderate accuracies were observed in multi-class systems attempting finer discrimination of force and speed levels.
Conclusions:
- This study presents a novel paradigm for establishing fNIRS-BCI systems.
- The findings suggest potential for increased degrees of freedom in future BCI designs using fNIRS.
