Wavelet-based method for removing global physiological noise in functional near-infrared spectroscopy

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Functional near-infrared spectroscopy (fNIRS) is a key non-invasive brain imaging technique.
• Global physiological noise significantly impacts fNIRS signal quality and research outcomes.
• Existing noise removal methods often require additional hardware or subjective analysis.

Purpose of the Study:

• To develop and validate a novel, data-driven wavelet-based method for removing global physiological noise from fNIRS signals.
• To improve the accuracy and spatial specificity of fNIRS data analysis.
• To offer a hardware-independent solution for fNIRS noise reduction.

Main Methods:

• A two-step wavelet-based approach was developed for noise removal.
• Wavelet transform coherence was used to identify time-frequency points affected by physiological noise.
• Wavelet energy suppression was applied to contaminated components, followed by signal reconstruction.

Main Results:

• The proposed method effectively removed global physiological noise in both simulated and real fNIRS data.
• Validation was performed on both task-based and resting-state fNIRS recordings.
• The method demonstrated improved spatial specificity for task activation and resting-state functional connectivity.

Conclusions:

• The novel wavelet-based method offers an effective solution for global physiological noise in fNIRS.
• This data-driven technique enhances the reliability and precision of fNIRS brain imaging.
• The approach holds promise for advancing cognitive neuroscience, clinical research, and neural engineering applications.