Initial-Dip Existence and Estimation in Relation to DPF and Data Drift

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Functional neuro-imaging relies on detecting brain activity.
• Early de-oxygenation, or initial dip, is a key indicator of cortical activity.
• Estimating the initial dip is crucial but challenged by data drift and differential pathlength factor (DPF).

Purpose of the Study:

• To investigate the existence and estimation of the initial dip in functional near-infrared spectroscopy (fNIRS) data.
• To analyze the influence of DPF and data drift on initial dip detection.
• To develop and validate a novel algorithm for drift estimation and hemodynamic response function (HRF) fitting.

Main Methods:

• Proposed an efficient algorithm for estimating drift in fNIRS data.
• Analyzed the effect of DPF on initial dip across four scenarios.
• Developed a neuro-activation model with iterative optimization for HRF fitting.
• Validated the algorithm on simulated and real-world fNIRS datasets.

Main Results:

• Shifting fNIRS signals to a transformed coordinate system aids in accurate information inference.
• Characterized HRF for varying DPF across different NIR wavelengths.
• The proposed model effectively estimates drift and optimizes HRF fitting.
• Algorithm performance validated on simulated data and six healthy subjects.

Conclusions:

• The developed algorithm enhances the accuracy of initial dip detection in fNIRS.
• Understanding DPF effects is critical for precise interpretation of hemodynamic responses.
• The novel neuro-activation model offers improved analysis of brain activity via fNIRS.