Functional near-infrared spectroscopy (fNIRS) measures changes in the brain through optical processes that determine the levels of oxygen and maps neurological activity. fNIRS provides high spatial resolution in the cortex; however, because reaction in the brain is not immediate, fNIRS provides relatively low temporal resolution.
The fNIRS is portable, easy to set up, and does not cause harm or discomfort to the subject. fNIRS probes are affixed to the subject’s forehead. The subject is then usually presented with a variety of stimuli on a computer screen.
The Edmond J. Safra Center conducts fNIRS research together with its partners at Drexel University. Research is currently being conducted in reading, mathematics, and underlying skills using this equipment.
Our laboratories have been using a rather new technology, fNIRS, to study the inner workings of the brain. Since functional Near-Infrared Spectroscopy (fNIRS) is such a new technique, a description of the technology, its uses and advantages are described below.
Steps towards utilization of Near-Infrared Spectroscopy (fNIRS) in neurocognitive research:
The unique contribution of to the study of brain activation related to cognitive task performance – a brief overview of the state of the art:
During the past decade, the field of cognitive neuroscience has experienced a tremendous gain in knowledge due to the development of methodological approaches that give more direct insight into the functions of the human brain than the classical behavioral measurements. On the basis of the quantification of changes in electrophysiological, metabolic or blood-flow parameters related to certain experimental tasks, well-established brain imaging methods such as EEG, PET or fMRI have promoted the development of explanatory models that linked certain cognitive functions assumed to underlie task performance for specific experimental settings to observed patterns of brain activity.
However, each of the above-mentioned methods suffers from certain fundamental methodological drawbacks that limit their application to a narrower range of scientific enquiries. While EEG offers an excellent temporal decomposition of electrical activity of the brain, its spatial resolution is lacking even now that powerful current source detection algorithms exist. Localization of differential brain activity using fMRI and PET, on the other hand, is outstanding, while the temporal dynamics of brain functioning are basically lost to these methods. Additionally, both PET and fMRI are very cost-intensive and, thus, available to only a small number of researchers in the field. What is more, the use of both methods is further restricted in that specific subject populations cannot be tested due to their substantial invasiveness. Finally, the much-needed combination of EEG with brain imaging methods that offer a fine-grained spatial resolution got stuck in the early stages of methodological development. For the foreseeable future, no suitable and validated methodological approaches to data analysis will be available that allow for artifact-correction when EEG- and fMRI-data are collected concurrently.
The very recent advent of functional Near-Infrared Spectroscopy (fNIRS), a method that may overcome some of these obstacles, may thus have considerable impact on research in the field of cognitive neuroscience. Based on the very same physiological parameters exploited by fMRI, this optical imaging technology has several advantages when compared to magnetic imaging:
- First and foremost, due to its affordability, it may be available to many laboratories in the near future.
- In contrast to fMRI, there are no restrictions to using fNIRS with specific subject populations such as certain patient groups; and, what is more interesting especially for developmental cognitive neuroscience; fNIRS can be used for studies on children, infants and even toddlers.
- The collection of fNIRS-data can be readily combined with EEG-measurement, which allows for the integration of neurophysiological data with high temporal and spatial resolution.
- The issue of ecological validity of experimental results should be addressed. fMRI-measurement requires that the subjects be confined to an extremely noisy, narrow and overall very unnatural environment during the data collection. This challenges the validity and reliability of individual study results.
- fNIRS is comparably insensitive to contamination with movement-related artifacts.
- Finally, fNIRS allows for the measurement of changes in both oxygenated (O2Hb) and deoxygenated (HHb) hemoglobin which is an advantage, since the O2Hb/HHb-ratio offers additional insight into the specifics of neurocognitive functioning (Cope & Delpy, 1988; Jobsis, 1977; Strangman, Boas, & Sutton, 2002; Ayaz, et al., 2011; Sela et al., 2011; Rolfe, 2000;Heekeren et al., 1997; Hofmann et al., 2008; Izzetoglu, et al 2004, Izzetoglu,2007; Sakatani, 1999, Suto et al 2002).
Overall, fNIRS is a neuroimaging method that, while currently not widely used due to its novelty, will have a considerable impact on research in cognitive neuroscience in the near future.
Some current limitations of functional Near-Infrared Spectroscopy:
While the advantages of fNIRS are impressive, some problems and limitations of this method have to be mentioned. Being an imaging approach that is still in its infancy, considerable methodological challenges have to be faced and dealt with:
- First of all, no standard procedures exist with respect to the analysis if fNIRS-data. There are no well-established routines or valid and reliable tools for data processing, let alone ready-to-use software packages that allow for the standardized pre-processing, analysis or visualization of the data. At the moment, many interested researchers are discouraged from using fNIRS as each user has to actually implement his/her own tools for data processing.
- One major problem with fNIRS-data is that due to what is called neurovascular coupling, task-related changes in the fNIRS-signal have to be assumed to originate not only from cerebral, but also from peripheral sources (e.g. blood-vessels in the scalp). As long as no validated approaches to artifact-correction exist, the usability of fNIRS is seriously restricted.
- The spatial resolution of fNIRS-data is lower than what researchers are used to from fMRI. Co-registration of anatomical and fNIRS-data may prove to overcome this problem to a certain degree. However, the routines have yet to be established.
- The method of mounting the optic fibers on the subjects’ heads is still lacking. Since the distance between light emitters and detectors (i.e. optodes) is as crucial for signal quality, as is the proper contact between the optodes and the scalp, intensive work has to be invested in order to establish the optimal setting for data collection.
- Finally, in order to allow for a partial replacement of fMRI-measurement by the fNIRS, the method has to be evaluated by systematical comparisons of fMRI and fNIRS results for a wide range of standard settings and applications. As long as there is no extensive data base that demonstrates the similarity of results from optical and magnetic imaging for the most basic and replicated experimental paradigms, implementing more sophisticated experimental design seems to be perilous.
This brief overview of only the most vital problems that researchers interested in using this new technology currently face, clearly demonstrates the necessity to invest into the development of appropriate theoretical and methodological approaches to the measurement and analysis of fNIRS-data. Apart from the implementation and validation of methods for the segregation of event-related changes in the fNIRS-signal reflected by cerebral and extra-cerebral tissue, the development of user-friendly routines and tools for the processing and visualization of the data seems to be the most pressing area of research. Investing know-how and manpower for the accomplishment of the required next steps in the development fNIRS-methodology may give us, i.e. the applicants, a head-start when it comes to high-impact research in the field of cognitive neuroscience.
