fNIRS-based brain-computer interfaces: a review.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Brain-computer interfaces (BCIs) offer communication pathways for individuals with severe motor disabilities by translating brain activity into device control.
• Functional near-infrared spectroscopy (fNIRS) is a non-invasive neuroimaging technique increasingly utilized for BCI applications.

Purpose of the Study:

• This paper reviews key components of fNIRS-based BCIs, focusing on signal generation, noise reduction, feature extraction, and classification.
• The review aims to provide a comprehensive overview of current methodologies and future directions in fNIRS BCI research.

Main Methods:

• Brain signal generation tasks, including motor imagery and cognitive tasks, are discussed in relation to common brain areas like the motor and prefrontal cortex.
• Noise removal techniques, from basic band-pass filtering to advanced methods like adaptive filtering and independent component analysis (ICA), are examined.
• Feature extraction methods (mean, variance, peak value, etc.) and classification algorithms (Linear Discriminant Analysis, SVM, HMM) are reviewed for their efficacy in fNIRS BCIs.

Main Results:

• fNIRS is particularly advantageous for cognitive tasks in the prefrontal cortex due to its non-invasiveness and lack of hair interference.
• While band-pass filtering is common for noise removal, advanced techniques are needed for overlapping signal bands.
• Linear Discriminant Analysis (LDA) offers a simple yet effective classification method for fNIRS BCI data.

Conclusions:

• fNIRS-based BCIs show promise for various applications, including monitoring neuroplasticity post-rehabilitation and neurostimulation.
• Future advancements are anticipated with hybrid EEG-fNIRS systems, bundled probes, and improved detection of early hemodynamic responses (initial dips).