fNIRS has many advantages over other neuroimaging modalities. It allows to measure blood oxygenation level changes related to neural activation in a direct and non-invasive manner, with quick set-up times and high temporal resolution. However, fNIRS signals may often be corrupted by measurement noise and physiology-based systemic interference. Therefore, successfully extracting neuronal activity-related signals from your fNIRS data requires careful statistical analyses.
Here, we offer an overview of fNIRS Analysis topics, ranging from fundamentals and study design, to analysis software options and advanced fNIRS analysis in real-time or multi-modal environments.
Click on any of the links below to learn more.
Can't find what you're looking for? Get in touch with us to find out more!
Three aspects of the human body are fundamental to fNIRS neuroimaging. First, human tissue is relatively transparent to light in the near-infrared range, thus allowing photons to propagate. Second, hemoglobin has unique absorbing characteristics, allowing for oxygenation-dependent quantification of NIR light absorption. Finally, the brains demand for oxygen is altered by neuronal activation. NIRS allows to measure relative changes in cortical hemoglobin concentration, coupling this concentration to brain functioning is what designates fNIRS.
During a NIRx Workshop in 2017, Dr. Ted Huppert from the University of Pittsburgh extensively covered the fundamentals of fNIRS Analysis. In his lecture, Dr. Huppert describes how detected light intensities can be converted into hemoglobin concentration changes, and how these concentration changes can be projected on brain models, on the subject-level or for group-level statistics.
Defining the right research paradigm may perhaps be the most crucial factor determining the success of scientific experiments. By carefully consulting the existing literature, and reviewing previous studies, a right experimental design can be constructed. This allows for optimal answering of research question, thereby tremendously reducing workload when the analysis stage is reached.
In this lecture by Dr. Ted Huppert from the University of Pittsburgh, which was given at a Workshop hosted by NIRx Medial Technologies at UT Austin in October 2017, key considerations of fNIRS experimental design are presented, as well as how experimental parameters relate to fNIRS analysis once data is acquired.
The fNIRS community has rapidly grown over the past decades, resulting into sophisticated and user-friendly analysis tools, each offering unique qualities that may suit your research best.
Selected Analysis Software platforms are:
The NIRS Toolbox is one of the newest and most sophisticated Analysis Software for fNIRS data at the moment. This toolbox consists of a set of MATLAB based tools, including functions for signal processing, visualization, and statistics of fNIRS data. The NIRS Toolbox is built around an object-oriented framework of Matlab classes and namespaces. The NIRS Toolbox has an active userbase, and is constantly in a state of improvement. NIRx fNIRS data recorded by NIRStar 15.0 can be directly imported in the NIRS Toolbox.
Dr. Ted Huppert from the University of Pittsburgh gives a talk on fNIRS Analysis and his "NIRS Toolbox" for MATLAB, at a workshop hosted by NIRx Medical Technologies at UT Austin on October 5th, 2017.
Download NIRS Toolbox (requires MATLAB) from
Dr. Ted Huppert's BitBucket website.
HOMER2 is one of the oldest fNIRS Analysis Software Options out there. It consists of a set of MATLAB scripts, which have evolved since the early 1990s. HOMER2 is used to convert fNIRS data into maps of brain activation. The software has a familiar and user-friendly interface, and supports group analyses and re-configuration of the fNIRS analysis processing stream. Additionally, HOMER2 allows the user to integrate own algorithms into the processing stream. All processing functions can also be accessed at the script level, for additional flexibility. NIRStar 15.0 offers the functionality to directly export data into the HOMER2 analysis format. Alternatively, the user may choose to convert previous recorded data using a MATLAB script.
Download Homer2 from homer-fnirs.org.
nirsLAB is an in-house creation of NIRx Medical technologies, since it has been developed in cooperation with the Optical Tomography Group of prof. Randall Barbour at the SUNY Downstate Medical Center. nirsLAB is a versatile fNIRS data analysis environment, which aims to offer new users an introduction into studying time-varying near-infrared measurements of tissue.
nirsLAB offers several data inspection tools, as well as statistic parametric mapping (SPM) environment. The GUI-based tool box provide the resources to easily edit and explore the many neuro-response features. Moreover, nirsLAB offers individual and group level analysis, meaning it allows to quantitatively evaluate single and multi-subject findings in a variety of display formats.
Download nirsLAB from NITRC.
fNIRS is a promising candidate for Brain-Computer Interface investigations. Its great performance in acquiring cortical signals in the presence of muscle movements and the possibility of setting up measurements in realistic environments, make this neuroimaging technique ideal for patients with for example ALS. Real-time fNIRS analysis requires sophisticated algorithms, that (for example) dynamically adjust predicted activation patterns. The Webinar video below offer insights into online processing of fNIRS signals for BCI or Neurofeedback paradigms.
In this Webinar organized by InsideScientific and NIRx, Dr. Bettina Sorger and Dr. Ujwal Chaudhary show how they use fNIRS to read out voluntarily controlled brain states in real-time. They discuss how acquired data is processed online, and used to classify cortical activation patterns. This allows for a communication channel with completely locked-in ALS patients, who are otherwise isolated from the external world.
Many research institutes appreciate multi-modal applications with fNIRS. By gathering information from several modalities, measurements can be rendered more robust, resulting into an advanced understanding of human brain functioning. Typical combinations are fNIRS and EEG, Eye-Tracking or fMRI, but tDCS and TMS have also been applied to concurrently modulate brain activity.
Dr. Ted Huppert from the University of Pittsburgh gives a talk on how measuring the brain concurrently with several modalities affects fNIRS Analysis, during the NIRx Medical Technologies UT Austin Workshop in October 2017.