Functional near-infrared spectroscopy (fNIRS) is a technique that uses light to study what is happening inside living tissue, including the brain. In simple terms, we shine light into the tissue and draw conclusions from the light that comes back out. But before we can understand how fNIRS works, we need to answer a question that might seem almost too simple: can light actually travel through the body?

The answer, as you are about to discover, is: it depends entirely on which kind of light you use.

In this lesson, we will see where near-infrared light sits in the electromagnetic spectrum, why tissue has an Optical Window, and which molecules set its limits. We will then follow how this simple observation grew into a technique used in the brain, in muscle, and at the bedside.

1. Light and the Electromagnetic Spectrum

For the purposes of this course, we will treat light as electromagnetic radiation that travels as a wave. The distance between the peaks of those waves (the wavelength, typically denoted as λ) determines what kind of light it is and how it interacts with matter. Across the electromagnetic spectrum, wavelengths range from kilometers (radio waves) to far less than a nanometer (gamma rays). For the light used in fNIRS, it is convenient to measure wavelength in nanometers (nm): one nanometer is one billionth of a meter (\(10^{-9}\) m), far smaller than the width of a human hair.

The light we see with our eyes spans roughly 400 to 700 nm (violet at the short end, through blue, green, yellow, and orange, to red at the long end of ~700 nm). But visible light is only a tiny slice of a much larger continuum called the electromagnetic spectrum. Beyond red light, at wavelengths longer than 700 nm, lies infrared radiation: light we cannot see but can sometimes feel as warmth. Beyond infrared, wavelengths continue to grow through microwaves and radio waves. In the other direction, below violet, wavelengths shrink through ultraviolet, X-rays, and gamma rays, each carrying progressively more energy.

Near-infrared (NIR) light refers specifically to the region just beyond visible red, roughly 700 to 2500 nm. It is invisible to the naked eye, but it can be generated cheaply with small LEDs or laser diodes, and detected with standard optical sensors. These practical properties alone made it attractive to early researchers. But what truly set NIR light apart was what it does when it encounters biological tissue.

2. The Optical Window: A Gap in Tissue Absorption

Try this simple experiment: press a small torch against the palm of your hand in a dark room. You will see your fingers glow red from the inside. Bones, tendons, and blood vessels appear faintly as darker shadows within the illuminated tissue. What you are witnessing is light (specifically red and near-infrared light) penetrating several centimeters of biological material and scattering back out through the skin.

Figure 2. Near-infrared transillumination of the hand.

This does not happen with blue or green light. Hold that same torch against your hand and look at whether blue light penetrates; it doesn’t. It is absorbed almost immediately within the first millimeter of skin. And it does not happen with far-infrared light either, because water (which makes up roughly 70% of your body) absorbs those wavelengths strongly. So why does red and near-infrared light behave so differently?

The answer lies in how the main light-absorbing molecules in tissue interact with light at different wavelengths. A chromophore is simply any molecule that absorbs light; the word comes from the Greek for “color carrier.” In biological tissue, the four chromophores that matter most for our purposes are:

🪪 Oxy-hemoglobin (HbO₂) Hemoglobin is the oxygen-carrying protein in red blood cells. When it binds oxygen, it becomes oxy-hemoglobin. HbO₂ absorbs strongly below ~600 nm (which is why oxygenated blood looks bright red), and it is relatively transparent in the NIR window above 700 nm.	🪧 Deoxyhemoglobin (HbR) Hemoglobin without oxygen. Like HbO₂, it absorbs strongly below ~600 nm and much less in the NIR, but its absorption differs from that of HbO₂ across the NIR range (and this spectral difference is precisely what fNIRS exploits to track oxygenation changes).
💧 Water (H₂O) Makes up ~70–80% of tissue. Nearly transparent in the visible and NIR range up to about 950 nm, then becomes a very strong absorber. Defines the upper boundary of the Optical Window.	🌞 Lipids & Melanin Lipids (fats) are broadly transparent in the NIR. Melanin is a pigment present in skin and hair that absorbs strongly in the visible but less so in the NIR. Both influence how well light couples into tissue.

We can quantify how much a chromophore absorbs at each wavelength with an absorption spectrum. If you plot the absorption of all these chromophores together against wavelength, a remarkable gap appears: between roughly 650 and 950 nm, all of them are at their lowest combined absorption. Hemoglobin’s strong visible absorption has tapered off, and water hasn’t yet started its strong infrared absorption. This gap where tissue is as transparent as it ever gets is called the Optical Window.

Within this window, NIR light can travel several centimeters through tissue before being absorbed. For the brain, this means light shone onto the scalp can reach the outermost layers of the cortex and return to a detector placed a few centimeters away, carrying information about the tissue it passed through. That is the physical foundation of fNIRS.

It is worth noting that this window is defined by biology, not engineering. Every fNIRS device, from a simple two-channel clinical monitor to a high-density research system, has to work within the same broad region of low absorption (roughly 600–1000 nm), and most use wavelengths between about 650 and 950 nm. There is no way around it. The window is what it is.

3. The Birth of fNIRS

The Optical Window had been noticed by physiologists as early as the 1870s, when it was known that tissue had some transparency to red light, at the long-wavelength end of the visible spectrum. But for nearly a century, nobody had turned this observation into a practical measurement tool for living tissue. That changed in 1977, when a paper by Frans F. Jöbsis, a physiologist at Duke University, appeared in the journal Science.

What made Jöbsis’s contribution so significant was not just the demonstration itself but the conceptual leap it represented. He recognized that the Optical Window, combined with the fact that oxy- and deoxy-hemoglobin absorb light differently, created an opportunity to monitor tissue oxygenation continuously, safely, and without touching what you were measuring. No other technique at the time could offer all three of these properties simultaneously.

It is worth noting that Jöbsis was studying both the brain and the heart (the myocardium) simultaneously. From the very first paper, NIRS was understood as a technique for monitoring oxygenation in any accessible tissue, rather than a brain-specific tool. This broader scope would prove important in the decades that followed.

4. From the Lab to the Field

In the decade following Jöbsis’s paper, researchers rapidly expanded the range of tissues and questions they applied NIRS to. The same Optical Window that allows light to reach the cortex in cats also allows it to penetrate muscle, the neonatal skull, and peripheral tissues. Three application domains emerged in parallel during the 1980s and early 1990s, each contributing insights that would later benefit brain imaging.

💪 Muscle Physiology & Sport Science During exercise, muscles consume oxygen rapidly. NIRS can track this in real time by measuring hemoglobin oxygenation in the muscle directly beneath the probe. Britton Chance and colleagues in Philadelphia were pioneers here, publishing landmark studies in elite rowers in 1992. Today, wearable NIRS muscle oximeters are used in professional sport, rehabilitation, and exercise physiology.

👶 Neonatal & Clinical Monitoring Jöbsis’s own group at Duke made the first bedside measurements of cerebral oxygenation in sick preterm infants (Brazy et al., 1985), and David Delpy and colleagues at University College London turned NIRS into a quantitative neonatal monitor, a population for whom MRI and CT are extremely difficult to use. The thin skulls of infants make them ideal candidates, and bedside NIRS monitoring in neonatal intensive care units became one of the earliest clinical applications of the technology.

This parallel development matters for a reason that goes beyond historical curiosity. Many of the methodological insights that shaped modern fNIRS (e.g., how to model the path photons take through tissue, how to account for the fat layer between probe and target tissue, how to deal with motion artifacts) were first worked out by researchers studying muscle, not brain. Understanding fNIRS as a technique that spans these domains gives you a richer and more accurate picture of how it works and where its limitations come from.

5. fNIRS: Functional Brain Imaging with NIRS

Through the late 1980s and into the early 1990s, NIRS was primarily a clinical and physiological monitoring tool. Researchers were using it to track gross oxygenation in patients, newborns, and exercising athletes, but not yet to map which brain regions were active during thought and behavior. That step required a further conceptual jump: not just monitoring whether tissue was oxygenated, but using small, task-related changes in oxygenation to locate and study brain function. That jump came in 1993, when at least four independent groups published within a few months of each other.

🎍 Japan, 1993 Hoshi, Y. & Tamura, M. (1993) “Detection of dynamic changes in cerebral oxygenation coupled to neuronal function during mental work in man.” Neuroscience Letters, 150, 5–8. → doi	🎍 Germany, 1993 Villringer, A. et al. (1993) “Near infrared spectroscopy (NIRS): A new tool to study hemodynamic changes during activation of brain function in human adults.” Neuroscience Letters, 154, 101–104. → doi
🎍 USA, 1993 Chance, B. et al. (1993) “Cognition-activated low-frequency modulation of light absorption in human brain.” Proceedings of the National Academy of Sciences, 90, 3770–3774. → doi	🎍 Japan, 1993 Kato, T. et al. (1993) “Human visual cortical function during photic stimulation monitoring by means of near-infrared spectroscopy.” Journal of Cerebral Blood Flow & Metabolism, 13, 516–520. → doi

Together, these papers showed for the first time that NIRS could detect changes in cerebral hemoglobin oxygenation in healthy human adults during cognitive tasks and visual stimulation. The signal was small, the equipment was bulky, and there were only a handful of measurement points on the scalp. Despite all the limitations and difficulties at the time, the principle of functional NIRS (fNIRS) as a neuroimaging modality was established.

In the three decades since, the field has grown from those first proof-of-concept experiments to an international research community with thousands of publications per year. Systems have shrunk from laboratory benches to wireless caps weighing a few hundred grams. Studies have moved from carefully controlled lab tasks to naturalistic environments, social interactions, sports fields, and classrooms. The technique now reaches populations that other brain imaging methods struggle to access, such as newborns, toddlers, people with severe motor impairments, and patients in intensive care.

 

A general, brief overview of how the field developed can be described as below:

---