Processing pipeline for image reconstructed fNIRS analysis using both MRI templates and individual anatomy

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Signal Processing

Background:

• Functional near-infrared spectroscopy (fNIRS) generates channel-based data, which can be transformed into volumetric representations for improved spatial analysis.
• Managing spatial variance in fNIRS data is crucial for accurate interpretation, especially when dealing with optode location variability.
• Existing methods for fNIRS data processing lack integrated pipelines for comprehensive image reconstruction.

Purpose of the Study:

• To present an innovative, integrated pipeline for the image reconstruction of fNIRS data.
• To provide a user-friendly, code-accompanied pipeline for processing fNIRS data from raw measurements to group-level analysis.
• To enable image reconstruction using either standardized MRI templates or individual subject anatomy.

Main Methods:

• Developed an integrated pipeline combining new and existing software tools for fNIRS data processing.
• Included steps for cleaning and preparing optode location data.
• Incorporated preparation and standardization of individual anatomical MRI images.
• Implemented a light model and 3D image reconstruction algorithms.
• Facilitated group-level data analysis in a standardized space.

Main Results:

• The pipeline was successfully tested using both MRI templates and individual anatomy.
• Validation was performed on fNIRS data from various collection systems.
• High temporal correlations were observed between channel-based and image-based fNIRS data.
• Reliability was demonstrated using a longitudinal study dataset of 74 children.
• Good correspondence was found between channel-space data and image-reconstructed data.

Conclusions:

• The presented pipeline offers a unique, integrated solution for fNIRS image reconstruction.
• This approach enhances the volumetric representation and spatial accuracy of fNIRS data.
• The pipeline facilitates a more comprehensive analysis workflow, from raw data to group analysis.
• Identified areas for future development in fNIRS data processing software.