How can a clip on a fingertip estimate oxygen in blood—and why might the same device give different accuracy across skin tones? A pulse oximeter sends red and near-infrared light through tissue and measures what reaches a detector. Its calculation depends on how blood absorbs those wavelengths. Skin is part of the optical path too, so its pigment can change the signal before the instrument ever analyzes it.
A ratio built from light
Oxygenated and deoxygenated hemoglobin, the oxygen-carrying protein in red blood cells, absorb red and near-infrared light differently. A pulse oximeter compares the changing absorption at these wavelengths to estimate oxygen saturation, the share of hemoglobin carrying oxygen. The pulse component helps distinguish arterial blood, which pulses with each heartbeat, from more constant absorption by other tissue. [1]
This is an indirect measurement: the device does not sample blood or count oxygen molecules. It relies on a model linking light signals to saturation. The model works best when the device receives a usable signal and when its assumptions fit the person and conditions being measured. That distinction matters because a neat numerical display can hide uncertainty in the optical path.
Pigment changes the path
Melanin, the pigment that gives skin much of its color, absorbs some red and near-infrared light. If less of the emitted light reaches the detector, the balance used in the calculation can shift. NHLBI describes pulse oximeters as tending to overestimate oxygen saturation in patients with darker skin tones, a concern that can be most consequential when oxygen is low. [1]
Imagine measuring a dim lamp through two tinted panes instead of one. The detector sees a combined effect, and a calculation calibrated for a different optical path may misread it. Skin tone is not the only influence: poor fingertip blood flow, movement, nail polish, and falling oxygen saturation can also make readings less reliable. [1]
A sensor is not a diagnosis
The broader class of biomedical sensors detects physical, chemical, or biological signals and turns them into data. A wearable can track heart rate or sleep patterns, for example, but tracking a signal does not by itself establish a diagnosis. NIBIB notes that many advanced sensors remain research tools rather than practical routine-care devices. [2]
For pulse oximetry, the distinction between a reading and the underlying condition matters. A displayed estimate is useful information for a patient or clinician, but it is not a direct view of all oxygen-related processes in the body. NHLBI also describes how inaccurate readings can delay care when an apparently reassuring value masks desaturation, so device performance affects decisions as well as numbers. [1]
Changing the measurement
One research approach described by NHLBI uses green light and reflected rather than transmitted light. Green light is absorbed differently by oxygenated and deoxygenated blood and does not penetrate as deeply through a finger. Reflection measurement may avoid some of the deeper tissue and pigment layers that complicate transmission measurements. [1]
The researchers also calibrated their device using a skin-pigment color scale, an attempt to account for variation rather than assume one optical path fits everyone. Wireless sensors that can collect data during daily activity are another development goal, but a prototype or research project is not proof of clinical performance. Better sensor design depends on testing across diverse people and settings, and on showing that improved readings support sound decisions. [1]
Explore the physics: an explanatory diagram
Sources & further reading
- Updating Pulse Oximeters NHLBI
- Sensors NIBIB
Prepared for the independent PSI Center reading library. About edition dates and source checking.
This explainer describes scientific principles. It does not assess individual treatment choices.
