NeuroDOT: a Matlab and Python Toolbox for Optical Brain Mapping
Poster No:
2258
Submission Type:
Abstract Submission
Authors:
Emma Speh1, Yash Thacker1, Ari Segel1, Dan Marcus2, Muriah Wheelock3, Adam Eggebrecht1
Institutions:
1Washington University School of Medicine, St. Louis, MO, 2Washington University in St. Louis, St. Louis, MO, 3Washington University School of Medicine in St. Louis, St. Louis, MO
First Author:
Co-Author(s):
Introduction:
Processing pipelines for functional near infrared spectroscopy (fNIRS) and diffuse optical tomography (DOT) data pose many challenges. Costly software licenses (e.g., Matlab) and a tendency to develop lab-specific pipelines prevent the adoption of standardized practices for data processing. Another challenge of fNIRS and DOT toolbox development is the need for compatibility with multiple data formats (e.g.,SNIRF, BIDS, NIRS, NIfTI, GIFTI, 4dfp, etc.) and a wide variety of commercially available systems (e.g., GowerLabs, NIRx, etc.). We are developing NeuroDOT as a Matlab and Python-based toolbox for fNIRS and DOT with functions and pipelines for data pre-processing, anatomical light modeling, image reconstruction, analysis, and visualization. Compatibility with various common data formats and the use of Python support community adoption of NeuroDOT alongside the modernization and expansion of NeuroDOT's tools with best software practices.
Methods:
NeuroDOT is a self-contained toolbox that contains individual functions and scripts for data processing. NeuroDOT provides tools for pre-processing, data quality analysis, head modeling, image reconstruction, and post processing with statistical analyses, removing the need to combine multiple packages to perform these tasks (Figure 1). NeuroDOT contains functions for the conversion between various data formats including volumetric data in the Neuroimaging Informatics Technology Initiative (NIFTI) and 4-dimensional floating point (4dfp) formats as well as converters between NeuroDOT's format, the Shared Near Infrared Spectroscopy Format (SNIRF), the Near Infrared Spectroscopy (NIRS) format, and Brain Imaging Data Structure (BIDS) specifications. Individual functions are organized into categories based on shared functionality: Analysis, Reconstruction, Temporal Transforms, Visualizations, and others. Documentation for efficient training for the novice and expert user is provided in the form of tutorial PowerPoints and Jupyter notebooks accompanying Matlab and Python scripts, respectively. XNAT, an open-source cloud-based platform, is utilized to develop containers for each of the NeuroDOT pipelines.
Results:
NeuroDOT is available for download on GitHub (https://github.com/WUSTL-ORL/NeuroDOT) and NITRC (www.nitrc.org/projects/neurodot). All functions involved in the NeuroDOT pre-processing and reconstruction pipelines have been refactored to Python. Pre-processing, image reconstruction, and full data processing scripts have also been converted to Jupyter notebooks in Python and have been deployed as containers on XNAT. A guide for the use of the NeuroDOT pipelines on XNAT is also provided. Shared datasets including resting state DOT data, DOT data recorded during retinotopy, and common hierarchical language paradigms, and DOT data recorded during language tasks including silent reading of single words and covert and overt single verb production are available to download from NITRC.
Conclusions:
Herein, NeuroDOT supplies the fNIRS community with a highly efficient and effective toolbox with shared extensible tools for fMRI-comparable high fidelity optical brain mapping. Functions are written as building blocks for workflows, so groups can build their own pipelines. We have aimed to promote modernization of the growing components of NeuroDOT by refactoring the toolbox in Python with enhanced development tools, shared data repositories, and data format standardization relevant to both optical and fMRI fields and have enhanced the support for our community of users and developers with expanded documentation and tutorials, detailed in Figure 2. Open-source software is essential for standardization of processing and accessibility of the NeuroDOT toolbox to new users. Next steps for this project include launching Optical-imaging XNAT-enabled Informatics (OXI) an XNAT-based platform for worldwide fNIRS and DOT data sharing and standardized or customized container-based processing on the cloud.
Neuroinformatics and Data Sharing:
Workflows 1
Novel Imaging Acquisition Methods:
NIRS 2
Keywords:
Data analysis
Open-Source Software
Workflows
Other - Diffuse Optical Tomography (DOT)
1|2Indicates the priority used for review
Provide references using author date format
Eggebrecht, A.T. (2014), “Mapping distributed brain function and networks with diffuse optical tomography,” Nature Photonics, 8, 448-454.
Ferradal, S.L. (2014), “Atlas-based head modeling and spatial normalization for high-density diffuse optical tomography: in vivo validation against fMRI,” NeuroImage, 85, 117-126.
Hassanpour, M.S. (2014), “Statistical analysis of high density diffuse optical tomography,” NeuroImage, 85, 104-116.
Schroeder, M. L. (2023). "Mapping cortical activations underlying covert and overt language production using high-density diffuse optical tomography." Neuroimage 276: 120190.
Sherafati, A. (2020), “Global motion detection and censoring in high-density diffuse optical tomography,” Human Brain Mapping, 41, 4093-4112.
White, B.R. (2009), “Phase-encoded retinotopy as an evaluation of diffuse optical neuroimaging,” NeuroImage, 49, 568-577.
Zeff, B.W. (2007), “Retinotopic mapping of adult human visual cortex with high-density diffuse optical tomography,” Proceedings of the National Academy of Sciences of the United States of America