Towards a standard analysis for functional near-infrared imaging

Area of Science:

• Neuroscience
• Biomedical Engineering
• Optical Imaging

Background:

• Functional near-infrared spectroscopy (fNIRS) measures brain activity via hemodynamic changes.
• A standardized analysis approach for fNIRS data is currently lacking, hindering its widespread application.
• Establishing robust analytical methods is essential for advancing optical imaging in neuroscience.

Purpose of the Study:

• To evaluate the general linear model (GLM) and frequency-domain analysis for fNIRS data.
• To establish standardized statistical approaches for analyzing optical imaging data.
• To investigate brain activation patterns during visual tasks using fNIRS.

Main Methods:

• Application of the general linear model (GLM) to fNIRS optical imaging data.
• Analysis of fNIRS data in the frequency domain using spatially resolved spectral analysis.
• Utilizing two visual paradigms: a checkerboard task and a moving colored stimuli task.

Main Results:

• GLM analysis successfully identified activation in the primary and secondary visual cortex during the checkerboard paradigm.
• A second activation focus, likely V5, was detected in the motion area during the moving stimuli task.
• Spatially resolved spectral analysis confirmed activation sites and revealed a delayed hemodynamic response in the motion area.

Conclusions:

• The general linear model (GLM) and spatially resolved spectral analysis are proposed as standard statistical methods for fNIRS.
• These methods are effective for detecting brain activation and characterizing hemodynamic responses.
• The proposed approaches are largely independent of differential path length factor assumptions, enhancing their applicability.