A Parametric Model for Characterizing Time-Variant Single Trials of Block-Design fNIRS Experiments

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Functional near-infrared spectroscopy (fNIRS) commonly uses block-design paradigms.
• Traditional analysis methods like GLM and WA assume a time-invariant brain system, which is often inaccurate.
• This assumption limits the understanding of dynamic brain responses during experiments.

Purpose of the Study:

• To develop and validate a parametric Gaussian model for quantifying time-variant brain activity in block-design fNIRS.
• To address the limitations of traditional time-invariant analysis techniques in fNIRS.
• To enable the study of dynamic, trial-to-trial changes in brain responses.

Main Methods:

• Proposed a parametric Gaussian model to quantify time-variant behavior in fNIRS data.
• Validated the model using simulated data across various signal-to-noise ratios (SNRs).
• Applied the model to recorded data from an auditory block-design fNIRS experiment.

Main Results:

• The proposed model successfully characterized Gaussian-like fNIRS signal features at SNRs ≥3dB.
• Analysis of recorded data revealed statistically significant, quantitative changes in fNIRS responses across trials.
• Model parameter values aligned with visual inspection of individual trial data, confirming its effectiveness.

Conclusions:

• The parametric Gaussian model effectively captures trial-to-trial differences in fNIRS responses.
• This technique allows for the investigation of time-variant brain activity within block-design fNIRS studies.
• Researchers can now better study dynamic neural processes using established block-design fNIRS paradigms.