19.12.2025 Usually, to examine the brain, we need a large MRI scanner and complex equipment. Researchers at ICTER suggest a different method: measuring brain activity from the subtle “noise” of scattered light in tissues.
The brain has no energy reserves. To function, it must be supplied constantly with oxygen and glucose through blood flow. Every activity – such as reading, solving a problem, or having a conversation – causes a local increase in blood flow. This dynamic change forms the basis of modern neuroimaging: instead of directly tracking neurons, we observe changes in blood supply.
Functional magnetic resonance imaging (fMRI) remains the standard in scientific research. It offers excellent spatial resolution, but it is expensive, loud, requires large equipment, and is not suitable for continuous monitoring of patients in intensive care units or during surgery.
These MRI limitations are why optical methods for “peeking” into the brain have been under development for years, involving techniques that use near-infrared light capable of penetrating the skull. Scientists from the Parallel Interferometry and Computational Optics (PICO) group at the International Centre for Translational Eye Research (ICTER) now propose interferometric speckle-contrast optical spectroscopy (iSCOS) as a potential solution. iSCOS is a method that combines the strengths of two previously competing approaches, and may bring us closer to convenient, non-invasive monitoring of cerebral blood flow. Details of this novel technique are described in the journal Biocybernetics and Biomedical Engineering.
From MRI to an optical “brain camera”
The best-known optical technique is near-infrared spectroscopy (NIRS). For the NIRS procedure, a patient wears a cap or headband equipped with LEDs and detectors. Light with a wavelength of approximately 700-900 nm penetrates the tissues, scatters, and partially returns to the detectors a few centimeters away. Based on changes in light intensity, it is possible to estimate the concentrations of oxygenated- and deoxygenated-hemoglobin, and indirectly assess brain activity.
Another important family of methods is diffuse correlation spectroscopy (DCS). Instead of brightness, DCS analyzes the “flicker” of light; i.e., tiny, ultra-fast intensity fluctuations that occur when photons scatter off moving red blood cells. From the decay rate of signal autocorrelation, one calculates a blood-flow index.
The problem is that classical DCS requires ultrafast detectors operating at megahertz frequencies. These detectors are expensive, have few channels, and generate enormous amounts of data. Parallel systems based on Single-Photon Avalanche Diode (SPAD) arrays perform better, but remain technologically complex.
What are speckles?
An alternative to tracking fast temporal fluctuations is analyzing spatial correlations; i.e., how “grainy” the image of scattered light in the tissue appears. If one illuminates the tissue with a coherent laser and captures the reflected light with a camera, a characteristic grainy pattern is observed, called speckles.
When blood flows faster, the speckle pattern blurs during the camera’s exposure. The more blurred it is, the lower the so-called speckle contrast. Simply stated, lower contrast means faster flow. This relationship is the basis of laser speckle contrast imaging (LSCI), which is widely used to image blood flow in the mouse brain or the retina.
Over the past few years, speckle-contrast optical spectroscopy (SCOS) methods have been developed to transfer the contrast-imaging concept to probe deeper into tissue, into the diffusive-scattering regime necessary for studying the human brain. Yet, two “worlds” continue to collide: classical DCS, which examines correlations in time, and speckle-based techniques, which analyze spatial blur.
Klaudia Nowacka-Pieszak has proposed something like a “bridge” between these approaches. Instead of choosing one, she uses instrumentation that allows both types of correlations to be computed from the same dataset.
iSCOS – what’s new in interferometry with an ultrafast camera?
The starting point for iSCOS is an earlier system developed at ICTER, CW-πNIRS, an interferometric setup in a Mach–Zehnder configuration with a 785-nm laser and an ultrafast CMOS camera. The laser light is split into a reference arm and a sample arm, which both go into the tissue. After passing through and scattering, both arms recombine, and an interferometric speckle pattern reaches the camera.
Two aspects are crucial. First, interferometry means that the optical field from the tissue is amplified by the reference field, increasing sensitivity to subtle changes caused by blood motion. Secondly, the camera is not a typical “slow” detector: it can record hundreds of thousands of frames per second, though on a small sensor area.
The brief narrow focus and ultra-rapid detector open an interesting possibility. From very short exposures (a few microseconds), it is possible to compute the full temporal autocorrelation function of the optical field, just like in DCS. Then, instead of asking the camera to lengthen exposures, one can synthetically “sum” sequences of frames and calculate what the speckle contrast would be for different exposure times. All of this happens in post-processing, from a single recording.
Klaudia Nowacka-Pieszak explains:
It is as if we had a raw high-frame-rate video and could freely adjust playback speed and exposure time without asking the camera to repeat the shot. This capability allows us to fairly compare methods based on temporal and spatial correlations – without ‘comparing apples to oranges’.
Interferometry has another advantage: it enables data acquisition at the level of the optical-field correlation function, g1, not only intensity. This ability simplifies theory, makes noise modelling easier, and opens the door to more sophisticated analyses.
Computer simulations and milk phantoms
The first stage of the research was conducted in silico. ICTER scientists generated virtual images, simulating scattering in tissue with flowing blood, and then gradually added artificial “noise” resembling real detector disturbances.
Using these data, they tested two ways of computing the signal. First, a simpler method relying only on speckle brightness was used; then, a more advanced method was used, which analyzes temporal changes before converting them into contrast. Under high noise levels, the simple approach begins to fail, producing falsely high estimates of blood flow, whereas the method using more complete temporal information better filters out disturbances and produces values closer to theoretical expectations.
In other words, iSCOS can not only examine temporal and spatial changes simultaneously, but also handles unavoidable measurement noise more robustly.
Next, actual tests were run, beyond virtual simulations. Instead of immediately studying humans, the team used optical phantoms; namely, solutions of milk in water, which scatter light similarly to tissue. They prepared samples of different “densities,” placed the probe at varying distances from the illumination point, computed signals using both approaches, and checked whether they produced consistent flow trends.
The results were encouraging. Both approaches to analyzing the signal showed similar trends, and speckle-contrast-based data were additionally more stable and less noisy, especially under difficult conditions with weak signals. This result suggests that iSCOS may be more resilient to disturbances and more suitable for applications requiring both sensitivity and repeatability.
First human test: reading text detected as blood-flow increase
The final step was a pilot in vivo test. A volunteer sat with their head stabilized, and a fiber-optic probe was placed on the forehead (3 cm separation). First, resting state was recorded, then activation was recorded during reading an unfamiliar text.
From the recorded data, the relative blood-flow index (rBFI) was computed separately from temporal autocorrelation and speckle contrast. During the reading task, rBFI increased by approximately 14% in the g1 analysis and approximately 32% in the speckle-contrast analysis. These values are similar to those reported by other parallel DCS systems based on SPAD arrays.
This demonstration is not yet a full clinical study; it is a verification that iSCOS genuinely “sees” activation of the frontal cortex. The number of channels was too small to clearly detect pulsation, and the experiment was performed on a single person with a single task. But as a prototype test, the results are quite promising.
Klaudia Nowacka-Pieszak says:
What excites us most is that within a single, relatively simple setup, we demonstrated consistency across theory, phantom measurements, and the first human trial. This suggests that the interferometric approach is not only feasible on paper but also under real measurement conditions.
What’s next? From the lab to the patient’s bedside
A long journey to clinical practice lies ahead. The current prototype requires a large ultrafast camera, a powerful computer and advanced data processing. Work is underway on continuous (“streaming”) acquisition and real-time analysis without memory limitations, as well as on miniaturizing the system, with smaller cameras, compact light sources, and more parallel channels.
Plans also include measurements at other wavelengths that penetrate tissue more effectively and return more photons from deeper brain layers. The literature also describes configurations combining such systems with ultrasensitive superconducting detectors (SNSPDs), which would further improve the signal-to-noise ratio.
In a broader research context, iSCOS will be linked with advanced retinal imaging methods developed at ICTER, such as spatio-temporal optical coherence tomography (STOC-T). This combined approach would enable the simultaneous study of hemodynamics in both the brain and the retina. Clinically, the goal is continuous, non-invasive monitoring of cerebral blood flow in patients with severe head trauma, after stroke, during cardiac surgery, in intensive-care units, or in functional studies of individuals who cannot lie inside an MRI scanner.
Klaudia Nowacka-Pieszak, Saeed Samaei, Dawid Borycki (2025). Interferometric speckle contrast optical spectroscopy (iSCOS) in continuous-wave parallel interferometric near-infrared spectroscopy (CW-πNIRS). Biocybernetics and Biomedical Engineering.
DOI: https://doi.org/10.1016/j.bbe.2025.11.001
Author: Scientific Editor Marcin Powęska
