Diffuse Optical Mapping of Human Brain Function using NeuroDOT and OXI: A Hands-on course
Instructors: Emma Speh, Adam Eggebrecht, Ari Segel
About this course
This course will teach modeling and processing approaches for using diffuse optical methods for mapping brain function in humans. This includes temporal data processing and image reconstruction using the NeuroDOT software package. This course will use a combination of instructor lecturing and hands-on exercises using publicly available high-fidelity data to teach both conceptual and practical aspects of fNIRS imaging using the NeuroDOT software package and Optical-imaging XNAT-enabled Informatics (OXI). Attendees will learn and develop skills using NeuroDOT pipelines including data pre-processing, anatomical light modeling, and image reconstruction. All the training will be centered upon providing knowledge, training and tools for volumetric imaging of functional brain activations based on contrasts of hemoglobin concentration as measured using fNIRS data. The class will review the basic physics and biology of the approach, step through how the software works locally and on the cloud, and train attendees to use the software through exercises. More information about NeuroDOT is available at https://www.nitrc.org/projects/neurodot.
What will you learn?
The format of the workshop will be a combination of lecture-style presentations and hands-on activities. During the lectures, participants will also be encouraged to ask questions to facilitate discussion. In the hands-on component, attendees will perform data pre-processing, data quality assessment, image reconstruction, array alignment, and head modeling using NeuroDOT in Matlab. Jupyter notebooks in Python for NeuroDOT pre-processing and image reconstruction pipelines will also be available for the attendees to use. Participants who participate in the hands-on tutorials will be able to perform a range of modeling and analyses using open-source tools in Matlab and Python. Attendees of the workshop will be able to interact directly with the presenters who are the main developers of NeuroDOT and OXI. Proposed ideas and feedback from the attendees will be used to guide future development of the NeuroDOT and OXI tools.
Course details
Course duration
3 hours
Prerequisites and/or other notes
Each participant will need access to their own laptop to participate in the hands-on sections. Matlab 2024b or newer, NeuroDOT, and NIRFASTer toolboxes should be installed beforehand. Required Matlab add-on toolboxes include o Signal Processing Toolbox o Deep Learning Toolbox o Image Processing Toolbox o Statistics and Machine Learning Toolbox o Parallel Computing Toolbox o Curve Fitting Toolbox
Delivery plan
Presentation 1: Intro to tissue optics fNIRS/DOT (20 min lecture, Adam Eggebrecht)
Presentation 2: System designs, data collection, and quality control (20 min lecture, Ari Segel)
Presentation 3: Data preprocessing and temporal analyses (20 min lecture, Emma Speh)
Presentation 4: Anatomical head modeling using AlignMe (20 min lecture + 30 min hands-on, Ari Segel)
Presentation 5: Data visualization with NeuroDOT (20 min lecture + 30 min hands-on, Ari Segel)
Presentation 6: Cross-lab Informatics with OXI (20 min lecture, Emma Speh)
Why should you enroll in this course?
The goal of the workshop is for attendees to understand current fNIRS and HD-DOT methods and their applications, quantitatively assess and track data quality, and utilize the NeuroDOT toolbox to perform data pre-processing, image reconstruction, and anatomical light modeling in a reproducible manner. Participants will be able to use OXI to test NeuroDOT pipelines on provided sample datasets, which they can then apply to their own data. Workshop attendees will learn about how to integrate data from different types of fNIRS systems, including custom and commercially available systems (GowerLabs LUMO, NIRx, Artinis, and more).