Interpolated functional manifold for functional near-infrared spectroscopy analysis at group level

Area of Science:

• Neuroscience
• Biomedical Engineering
• Data Science

Background:

• Group-level analysis of functional Near-Infrared Spectroscopy (fNIRS) connectivity can be cumbersome with classical methods.
• Manifold-based approaches offer superior data exploration but lack systematic surface selection.
• Infinite surfaces can intersect data point clouds, complicating groupwise analysis.

Purpose of the Study:

• To introduce a systematic method for selecting surfaces in manifold-based group-level connectivity analysis.
• To minimize surface interpolation error in functional Near-Infrared Spectroscopy (fNIRS) group analysis.
• To enhance the explorative capabilities of group-level fNIRS connectivity analysis.

Main Methods:

• Developed the Interpolated Functional Manifold (IFM) using radial basis functions (RBF) to model changes in hemoglobin concentrations.
• Evaluated root mean square error (RMSE) for four RBF families.
• Validated the IFM model against psychophysiological interactions (PPI) analysis using the Jaccard index (JI).

Main Results:

• Achieved lowest interpolation RMSE values of [value] for [units] and [value] for [units].
• Demonstrated strong agreement with classical group analysis ([value]) and PPI analysis ([value] and [value]).
• Successfully decoded group differences using ANOVA analysis ([value], [value], [value]).

Conclusions:

• The Interpolated Functional Manifold (IFM) provides a practical solution for selecting appropriate manifolds in group-level fNIRS data.
• IFM facilitates the application of manifold-based methods for enhanced group analysis of fNIRS datasets.
• This approach improves the systematic exploration of connectivity in neuroimaging studies.