Blood, light, and a matter of time. ICTER investigates circulation

Blood, light, and a matter of time. ICTER investigates circulation

How can we efficiently collect light returning from tissues while also identifying what it tells us about blood flow in the deeper tissues? Conventional interferometric near-infrared spectroscopy (iNIRS) can distinguish signals according to the amount of time that light takes to travel through tissue, but it collects only a few photons. Camera-based systems capture many more photons, but this distinction is not realized when living tissue changes rapidly. Researchers associated with ICTER combined both approaches in a hybrid system. They showed that an appropriately analyzed camera signal can reproduce the blood-flow response observed with iNIRS for late-arriving photons, which are more sensitive to deeper tissues.

When blood flow is measured optically at the forehead, the detector receives mixed information. Some comes from circulation in the scalp, while some comes from deeper tissues. This makes it difficult to determine what an observed change really tells us about the brain's blood supply.

The problem was investigated by Marcin Marzejon, PhD, Eng., who previously worked at the International Centre for Translational Eye Research (ICTER) at the Institute of Physical Chemistry of the Polish Academy of Sciences, and Dawid Borycki, PhD, DSc, leader of the PICO group. Their paper "Hybrid interferometric near infrared spectroscopy (hybrid iNIRS) enables time-of-flight–resolved non-invasive blood flow monitoring in humans in vivo" was published in "Scientific Reports".

The authors built a hybrid system that enables information about blood-cell motion to be considered during the time it takes for light to travel through tissue. They then tested it on the forearm and forehead. They showed that separating light according to its time of flight helps increase the contribution of information from deeper regions. This article describes a demonstration of the method's capabilities involving two healthy volunteers.

What do blood cells do to light?

The study used a laser with a wavelength of approximately 780 nm, in the near-infrared range. The light passing through tissue is scattered repeatedly. At the detector, the exiting light forms a grainy pattern of bright and dark spots, known as speckles. The motion of red blood cells causes this pattern to change. Analyzing the rate of changes in the speckle pattern makes it possible to calculate a blood-flow index, which is a measure of the dynamics of circulation in the tissue being examined.

Time of flight provides a different kind of information. Photons, or packets of light, that arrive later at the detector have usually travelled along longer paths. Their signal contains a greater contribution from deeper tissues. By selecting groups of photons with different times of flight, researchers can vary the measurement's depth sensitivity. However, each group still includes light from many paths passing through both superficial and deeper regions.

"Time of flight helps us select light with a greater contribution from deeper tissues. When we combine this information with changes caused by blood-cell motion, we can better distinguish signals from superficial circulation from blood flow in deeper tissues," says Dawid Borycki, PhD, DSc.

The cuff stops inflow; the headband tests the surface

The study involved two healthy men aged 33 and 43. Light was delivered into the tissue and collected through optical fibers positioned 1-cm apart. The forearm experiment compared three conditions: before inflation of a blood pressure cuff, during the blockage of blood inflow, and after the pressure was released. Each recording lasted one second.

After the cuff was inflated, changes in the light pattern slowed markedly, corresponding to decreased blood-cell motion. Once the pressure was released, the changes accelerated. This scenario is analogous to what would occur in reactive hyperemia; i.e., a temporary increase in blood supply following a period in which it had been blocked. Groups of late-arriving photons produced a larger response after cuff release than early-arriving photon groups, consistent with greater sensitivity to blood flow in deeper tissues.

In the forehead experiment, an elastic headband was used. Tightening the band restricted blood flow in the scalp. Because of the skull's rigidity, the researchers expected circulation inside it to remain largely unchanged. Signals with shorter and intermediate times of flight responded differently from late-arriving signals. The dynamics of the latter were better preserved, consistent with a greater contribution from deeper tissues. This experiment showed that time of flight helps identify the influence of superficial circulation. It does not, however, independently confirm blood flow in a specific region of the brain. Each group of photons still contains a mixture of information from different depths.

"By tightening the headband, we were changing circulation primarily in the scalp. Comparing the responses of different groups of photons allowed us to assess how strongly it influences the measurement. Late-arriving photons were less sensitive to this superficial compression," says Marcin Marzejon, PhD, Eng., the paper's first author.

After the headband was tightened, the calculated blood-flow index could increase even though superficial circulation was being restricted. The authors explain this result as a change in the proportions of the signals. The contribution of the more slowly changing signal from the scalp decreased, while deeper regions became more prominent; the index describes this mixture. Accordingly, an increase in the index cannot be taken as straightforward evidence that more blood had begun to flow in the brain.

One photon in a million

Very little light reaches the detector. According to the theoretical model for the configuration used by the authors, approximately one photon is collected for every million launched. The conventional iNIRS arm preserves information about the time of flight of light and changes caused by blood-cell motion, but its single-mode fiber collects only a few photons. The camera arm instead uses a multimode fiber and observes many light speckles simultaneously, collecting light more efficiently and providing better signal quality.

Each approach has a different limitation, however. In the hybrid system, the camera readout is about a hundred-times slower than acquisition in the conventional iNIRS arm. Therefore, the camera detector is not fast enough to distinguish times of flight while also tracking rapid signal changes in living tissue. In the human experiments, the camera consequently operated without time-of-flight discrimination, while conventional iNIRS provided a time-frame of reference showing how the response varies between different groups of photons. Combining the two arms made it possible to assess how much information about blood flow in deeper tissues can be obtained from the more efficient camera measurement. In both arms, weak light from tissue is combined with reference light, making it easier to detect as an interference in the reference light by the tissue emission, hence the term "interferometric." The method as a whole is called hybrid interferometric near-infrared spectroscopy (hybrid iNIRS).

Before the human experiments, the authors tested the system using an optical fiber and a mixture of milk and water. They used milk containing 3.2% fat, diluted at a ratio of 1:1 with water. This liquid mixture scatters light in a way similar to tissue, and allows the instrument to be fine-tuned under controlled conditions.

The preliminary tests helped establish the laser's operating speed. At rates of up to about 10,000 tuning cycles per second, the instrument could not keep up with rapid signal changes, especially for late-arriving photons. At around 100,000 cycles per second, fewer photons were available per cycle, and noise caused the calculated rate of change to be overestimated. For the human measurements, a rate of 50,000 cycles per second was selected as a compromise between speed and data quality.

The calculated time-of-flight resolution for these settings was approximately 95 picoseconds, or 95 trillionths of a second. The ability to detect such small differences in time helps to distinguish light returning along different paths.

How to prepare the camera?

Combining the two techniques made it possible to answer a practical question: when does a camera measurement, despite lacking time-of-flight information, produce a response similar to that obtained for late-arriving photons in iNIRS? The authors compared changes in the relative blood-flow index after cuff release. With the camera data, they varied the range of time lags included in the signal-correlation analysis.

When the analysis included shorter time lags, the camera response was close to that of late-arriving photons, which are more sensitive to blood flow in deeper tissues. Shifting the analysis towards longer time lags produced a response resembling that of earlier-arriving photons. Conventional iNIRS thus provided a reference for selecting how to analyze the more efficient camera signal.

This is one of the paper's most important findings: the hybrid system showed under what conditions an appropriately processed camera measurement can reproduce a selected response of late-arriving photons. This opens up the possibility of calibrating future devices' sensitivity to blood flow in deeper tissues by choosing how the data are analyzed.

How could patients benefit?

The aim of further development is non-invasive monitoring of blood supply to the brain and peripheral tissues, eventually also using portable or wearable devices. Such measurements could provide additional information when monitoring patients with head injuries, stroke, or a risk of hypoxia. The prototype has not yet been tested with such groups.

Two healthy participants are too few to assess differences between patients and the device's robustness to motion. In the human experiments, the camera operated without time-of-flight discrimination, and measurements from the two arms were performed sequentially. The authors plan to improve data acquisition and processing, and to use several wavelengths to distinguish the forms of hemoglobin as a measure of blood oxygenation.

The paper's greatest achievement is to show how efficient light collection by a camera can be combined with the time-of-flight information available in conventional iNIRS. Using the hybrid system, the researchers established that an appropriately analyzed camera signal reproduces changes in the blood-flow index observed for late-arriving photons, which are more sensitive to blood flow in deeper tissues. In this way, iNIRS becomes a reference for selecting how camera measurements are analyzed. This opens the way to develop practical devices whose sensitivity to deeper tissues can be adjusted through data processing.

Marcin Marzejon, Dawid Borycki (2026). Hybrid interferometric near infrared spectroscopy (hybrid iNIRS) enables time-of-flight–resolved non-invasive blood flow monitoring in humans in vivo. Scientific Reports

Author: Scientific Editor Marcin Powęska