Oxygenation, Flow, and Metabolism: What fNIRS Actually Measures

In the previous two lessons, we learned that fNIRS measures changes in oxy-hemoglobin (HbO₂) and deoxy-hemoglobin (HbR) by exploiting the different ways these two molecules absorb near-infrared light. You might reasonably assume that measuring oxygenation gives you a direct window onto what the tissue is doing: more HbO₂ means more activity, more blood flow, or more metabolism.

Although it can often be correct, this assumption is among the most common misconceptions in fNIRS research and can lead to seriously flawed interpretations of fNIRS data. This lesson is here to correct it before it takes root.

NIRS measures oxygenation. But oxygenation is not a single physiological variable. It is the net result of a balance between two competing processes: oxygen delivery to tissue and oxygen consumption by tissue. Understanding this balance is essential for interpreting any fNIRS signal correctly.

1. The Oxygen Balance: Delivery vs. Consumption

Every cell in our body needs oxygen to produce energy. The oxygen it uses comes from the blood, delivered through an intricate network of arteries, capillaries, and veins. At any moment, the oxygenation state of the blood in a given tissue reflects the balance between the amount of oxygen arriving and the amount being consumed. Think of it like water in a tank: the level depends on both the flow of water in and the flow of water out.

The physiologist Adolf Fick formalized this balance in the 19th century. Fick’s principle states, in essence, that the amount of oxygen consumed by a tissue equals the difference between the amount arriving (oxygen delivery) and the amount leaving (venous oxygen content), multiplied by blood flow.

This has a direct consequence for fNIRS. When you observe a change in HbO₂ or HbR in your data, that change could have been caused by a change in blood flow, a change in oxygen consumption, a change in blood volume, or some combination of all three. The fNIRS signal itself does not tell you which. Understanding your data means reasoning about which of these physiological changes is most likely to cause it, given your experimental context.

2. Oxygenation is Not Flow

Perhaps the best way to see why oxygenation and flow are different is through a concrete example. Consider what happens to a working muscle during intense exercise, exactly the situation that Britton Chance and colleagues were studying in elite rowers in the early 1990s, as we discussed in Lesson 1.1.

During maximal exercise, blood flow to the working muscle increases dramatically, sometimes by a factor of 20 or more compared to rest. A naive assumption would be that more flow means more oxygen delivered, so oxygenation should go up. But what actually happens in the muscle during intense exercise is often the opposite: HbO₂ decreases and HbR increases. The muscle is consuming oxygen so rapidly that even the vastly increased flow cannot keep pace with demand. The net result is deoxygenation, despite very high blood flow.

This example vividly illustrates the core point: you cannot infer blood flow from oxygenation alone. The same principle applies in the brain, in the heart, and in any other tissue you might measure with (standard) fNIRS. The direction of the oxygenation change depends on which process (i.e., delivery or consumption) dominates at that moment.

3. Three Numbers, Three Different Things

When you read fNIRS papers or look at data, you will encounter several related but distinct quantities. It is important to understand what each one means and what it does (and does not) tell you.

ΔHbO₂ and ΔHbR: changes in concentration

The quantities measured by standard fNIRS are changes in the concentrations of HbO₂ and HbR over time, typically denoted ΔHbO₂ and ΔHbR. (The Greek letter delta (Δ) means “change in.”) These are relative changes from a baseline, not absolute concentrations. You know that HbO₂ went up or down by a certain amount compared to where it started, but you do not know the actual concentration present in the tissue at any point. This is a fundamental limitation of standard fNIRS.

These concentration changes reflect the combined influence of blood flow, blood volume, and oxygen consumption all simultaneously. An increase in ΔHbO₂ could mean more oxygenated blood arrived (increased flow), less oxygen was consumed (decreased metabolism), or more blood pooled in the tissue (increased volume) without a corresponding increase in consumption. You need additional information (or at least a well-reasoned experimental design) to distinguish between these possibilities.

Total hemoglobin (HbT): a proxy for blood volume

Total hemoglobin (HbT) is simply the sum of both forms: HbT = HbO₂ + HbR. Because the total number of hemoglobin molecules in a tissue volume depends on how much blood is present (regardless of its oxygenation state), HbT is generally interpreted as a proxy for local blood volume. When blood flow to a region increases, more red blood cells arrive, and HbT rises. When blood leaves a region, HbT falls.

This makes HbT a useful companion measure to ΔHbO₂ and ΔHbR. If HbO₂ increases while HbT also increases, the most likely interpretation is that more oxygenated blood has arrived in the tissue. If HbO₂ increases but HbT is unchanged, the more likely interpretation is a redistribution of the existing blood, i.e., the same hemoglobin molecules becoming more oxygenated because less oxygen is being consumed. As you can see, these two scenarios lead to the same oxygenation change, but very different physiological stories.

Oxygen saturation (StO₂)

Oxygen saturation (StO₂, or tissue oxygen saturation) is yet another quantity that expresses the fraction of total hemoglobin that is in the oxygenated form:

Oxygen saturation is an absolute quantity: it tells you the actual oxygenation state of the tissue at a given moment, not just how much it has changed from some arbitrary baseline. This makes it much more physiologically interpretable than ΔHbO₂ alone. A saturation of 60% means something specific; a ΔHbO₂ of +0.1 µM means the tissue is slightly more oxygenated than it was a moment ago, but you have no idea whether the baseline was 40% or 80%.

You are probably familiar with oxygen saturation from pulse oximetry, which reads SpO₂, the arterial oxygen saturation of peripheral blood. StO₂ in fNIRS is the tissue-level equivalent, averaging across the arterial, capillary, and venous blood in the measurement volume rather than sampling arterial blood alone.

4. Beyond Oxygenation: Different fNIRS Measures 

Now that you understand the basic physiological principles, it is worth being precise about what standard fNIRS (i.e., the most common NIRS technique, continuous wave (CW) fNIRS) actually gives you, and where its limits lie.

The reason CW fNIRS cannot provide absolute concentrations (and therefore cannot compute StO₂) comes back to the points we discussed in Lesson 1.2. Recall that to extract hemoglobin concentrations from light attenuation, you need to know how far photons actually travelled through the tissue. In CW systems, this path length is unknown and must be assumed. That assumption introduces an uncertainty that prevents recovery of absolute values; it only allows changes to be tracked reliably (more on that in the following Section).

However, the limitation of CW fNIRS is not an insurmountable one. It arises from the fact that a single intensity measurement cannot separate the contributions of absorption and scattering (and scattering is what determines path length). If you can measure something beyond just intensity, you can recover the missing information.

Two more specialised classes of NIRS devices can do exactly this:

Both TD and FD systems are more complex and expensive than CW systems, and they remain less common in research settings. But they represent a genuinely richer measurement: by recovering absolute hemoglobin concentrations, they allow you to compute StO₂ directly, to make meaningful comparisons of oxygenation levels between participants and sessions, and to get closer to estimating oxygen consumption.

A third class of NIRS techniques go one step further and measures blood flow directly. Diffuse Correlation Spectroscopy (DCS) and its newer variants – Speckle Contrast Optical Spectroscopy (SCOS) and interferometric DCS (iDCS) – use coherent near-infrared light to detect the random motion of red blood cells moving through tissue. Rather than measuring how much light is absorbed, these techniques measure how rapidly the detected light fluctuates over time: faster fluctuations indicate faster-moving red blood cells, which is a direct index of blood flow. Because they operate in the same Optical Window and use similar hardware to fNIRS, DCS systems can be combined with CW or TD/FD fNIRS in a single device, measuring oxygenation and flow simultaneously from the same tissue volume. This combination is particularly powerful: recall from Section 1 that estimating oxygen consumption requires knowing both flow and oxygenation together, therefore DCS-fNIRS hybrid systems are one of the most promising tools for non-invasive estimation of cerebral metabolic rate of oxygen (CMRO₂), the closest thing to a direct optical measure of brain metabolism. Like TD and FD systems, DCS devices are more specialised and less common than standard CW fNIRS; Unit 4 of this course will explain in detail how these device types work and what the trade-offs are.

Society for Functional Near-Infrared Spectroscopy.
Content is for educational purposes.