Surface-based integration approach for fNIRS-fMRI reliability assessment.

Area of Science:

• Neuroscience
• Biomedical Engineering

Background:

• Integrating functional near-infrared spectroscopy (fNIRS) and functional MRI (fMRI) is crucial for comprehensive brain activity analysis.
• Current integration methods lack standardization, leading to reproducibility issues in spatial agreement and temporal correlation assessments.

Purpose of the Study:

• To propose and validate a novel method for integrating fNIRS and fMRI data over the cortical surface.
• To enhance the reproducibility of spatial agreement and temporal correlation analyses between fNIRS and fMRI.

Main Methods:

• Developed a novel method projecting subject- and group-level source maps onto the cortical surface mesh.
• Defined anatomically constrained functional Regions of Interest (acfROI) for data integration.
• Quantified spatial agreement using Dice Coefficient and temporal correlation using Pearson's correlation coefficient.

Main Results:

• Subject-level analysis showed moderate to substantial spatial agreement (Dice Coefficient 0.43-0.64) for BOLD vs. HbO2.
• Group-level results demonstrated strong temporal correlation (0.95-0.98 for BOLD vs. HbO2; -0.91 to -0.94 for BOLD vs. HbR).
• Some instances showed a lack of spatial agreement for BOLD vs. HbR at the subject level.

Conclusions:

• The proposed method effectively integrates fNIRS and fMRI data by projecting source maps onto the cortical surface.
• Results indicate significant spatial and temporal correspondence between fNIRS and fMRI signals.
• This method supports the use of fNIRS in ecological settings, such as longitudinal brain activity monitoring.