Toward more intuitive brain-computer interfacing: classification of binary covert intentions using functional
Han-Jeong Hwang1, Han Choi2, Jeong-Youn Kim2
1Kumoh National Institute of Technology, Department of Medical IT Convergence Engineering, 61 Daehak-ro, Gumi, Gyeongbuk 730-701, Republic of Korea.
Journal of Biomedical Optics
|April 7, 2016
Summary
Directly decoding internal yes/no intentions using functional near-infrared spectroscopy (fNIRS) achieved 75% accuracy in healthy subjects. This brain-computer interface (BCI) method offers a more intuitive communication system for individuals with motor disabilities.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Human-Computer Interaction
Background:
- Traditional brain-computer interfaces (BCIs) often use arbitrary mental tasks for binary communication.
- A pilot study suggested direct intention decoding from functional near-infrared spectroscopy (fNIRS) for intuitive BCI communication.
Purpose of the Study:
- To investigate the reliability of fNIRS-based direct intention decoding for practical BCI communication.
- To evaluate the performance of classifying internal binary intentions using fNIRS.
Main Methods:
- Eight healthy subjects participated, answering 70 disjunctive questions.
- Brain hemodynamic responses were recorded using multichannel fNIRS while participants internally expressed 'yes' or 'no'.
- Various feature types, numbers, and time windows were tested for optimal classification.
Main Results:
- Approximately 75% of binary intentions were correctly classified using optimal feature sets.
- Oxygenated and deoxygenated hemoglobin responses showed high classification accuracy (75.89% ± 1.39 and 74.08% ± 2.87).
- The kurtosis feature yielded the highest mean classification accuracy; wide brain regions showed hemodynamic responses.
Conclusions:
- fNIRS-based direct intention decoding is a reliable method for BCI communication.
- This approach has the potential for more intuitive and user-friendly communication systems for patients with motor disabilities.
