Modern OCT devices can look inside the eye without a scalpel - but they do not always capture everything that matters. Some critical information is simply lost, blurred by scattered photons. The STOC-T technology, developed for years by Prof. Maciej Wojtkowski and his team at ICTER, aims to resolve this problem - not through post-processing, but by fundamentally changing the way imaging data is collected.
Optical coherence tomography (OCT) has become one of the foundations of modern ophthalmology. A patient sits in front of the device and focuses on a target, and moments later the physician can see a detailed cross-section of the retina - layer by layer, without any physical contact with the eye. It is one of the most significant advances in ophthalmology over the past three decades. OCT enables clinicians to detect glaucoma, age-related macular degeneration, diabetic retinopathy, and macular edema before a patient experiences noticeable vision loss.
Yet, behind this seemingly simple examination lies complex physics. Light entering the eye does not return to the detector in perfect order. Some photons carry useful information about the tissue being examined. Others become scattered, bounce in random directions, and mix with the desired signal. The result is an image with reduced contrast, increased graininess, and lower clarity. Ironically, it is often within these blurred details that the earliest signs of disease may be hidden. Prof. Maciej Wojtkowski, who heads the International Centre for Translational Eye Research (ICTER) at the Institute of Physical Chemistry within the Polish Academy of Sciences in Warsaw, has been working on this challenge for many years. In the article "Spatio-temporal optical coherence imaging and tomography for in vivo applications," published in the Journal of Biomedical Optics, he describes the STOC technology and its three-dimensional implementation, STOC-T. The method is designed not only to improve image quality but, more importantly, to separate true tissue signals from noise, already at the data-acquisition stage.
Why are eye images sometimes incomplete?
The retina is thin, complex, and delicate, and the choroid - the vascular layer located directly beneath it - is even more difficult to visualize at full resolution. Photoreceptors, the cells directly responsible for vision, are only a few micrometers in size. All of these characteristics of the components of the eye make imaging very challenging, especially considering the properties of the stimulating light.
When a photon originating from a single point in the eye tissue impinges upon multiple detector pixels instead of just the correct one, the resulting image is no longer a faithful representation of the structure being examined. Prof. Wojtkowski refers to this phenomenon as "optical crosstalk" (OC). For a physicist, OC is a matter of interference and loss of coherence . For a physician, OC results in a loss of diagnostic information. For a patient, OC may increase the risk that early pathological changes remain undetected.
"In imaging living tissues, the goal is not simply to collect as much light as possible. We also need to know which light tells us something meaningful about the tissue and which light merely degrades the image," says Prof. Maciej Wojtkowski.
How does STOC-T improve the image?
STOC-T is not another image-processing filter applied after signal acquisition. Instead, it changes the way data are collected. The system repeatedly modifies the phase of the light illuminating the tissue, using different spatial patterns known as phase masks. The recorded signals are then compared and averaged. Light that has been scattered behaves chaotically and differently for each mask, causing its contribution to the image to gradually cancel out. In contrast, light carrying reliable structural information remains stable and becomes increasingly dominant after averaging.
A useful analogy is trying to hear a specific conversation in a crowded, noisy room. If the background noise is random while the desired voice remains consistent, a well-designed recording system can isolate it. Likewise, STOC-T focuses on the desired signals, but with light instead of sound, and in real time.
"We do not treat noise as something that should later be cosmetically corrected. Instead, we try to design the measurement process in such a way that the interfering signals never have a chance to contaminate the image in the first place," adds Prof. Maciej Wojtkowski.
In highly scattering tissues, part of the information may be irretrievably lost at the moment of acquisition. Thus, no post-acquisition algorithm can fully reconstruct an accurate image from an already distorted image. STOC-T addresses the problem earlier - before the image is constructed.
What has the team demonstrated?
The publication presents results obtained both in laboratory models and in living biological tissues. In one experiment, Prof. Wojtkowski's team imaged a standard-resolution target covered first with a strongly scattering artificial layer, and then with a 100-micrometer-thick layer of rat skin, representing natural tissue. Without STOC-T, the image was heavily distorted in both cases. Once phase modulation was applied, the target structures became clearly visible again. The experiment illustrates the scale of the challenge: the object was always present, but without STOC-T the information about it was effectively lost on its way to the detector.
The most significant results, however, come from application of the technique to retinal imaging. STOC-T enables visualization of retinal layers, photoreceptors, ganglion cells, and the microstructure of the choroid, with a lateral resolution of approximately 5 micrometers - approaching the scale of individual cells.
Furthermore, STOC-T facilitates optoretinography (ORG), which records photoreceptor responses to light. Knowing what these light-sensing cells look like is advantageous, but knowing whether they function properly is often the more clinically relevant question. The journal article describes measurements of cone-photoreceptor responses to flickering light at frequencies between 1.5 and 45 Hz. The derived time constants were approximately 398 ms and 43 ms. These values closely match photoreceptor-activity measurements obtained using patch-clamp recordings in primate retinas, suggesting that STOC-T-based ORG may indeed reflect local cone responses. Such information is particularly valuable in diseases where cellular function begins to deteriorate before structural changes become visible. In these cases, the eye may still appear normal, while the cells are already behaving differently.
Who benefits from improved imaging?
According to the World Health Organization, at least 2.2 billion people worldwide live with visual impairment. In more than one billion cases, vision loss could have been prevented or could still be treated if diagnosed earlier and more accurately. With glaucoma, lost nerve fibers are difficult to restore. With diabetic retinopathy, vascular changes detected too late may lead to severe complications. With macular diseases, rapid diagnosis and precise treatment monitoring can determine whether a patient retains central vision.
STOC-T is not yet a clinical product. The report clearly outlines its current limitations. The method requires a high-speed CMOS camera (512 x 512 pixels operating at 60,000 frames per second) and a tunable laser covering the 800-870 nm wavelength range; and it generates enormous amounts of data. A single acquisition may exceed 8.5 GB, creating substantial computational challenges.
The researchers also see great potential in using multimode optical fibers as an alternative phase-modulation mechanism. For example, using a fiber with a 50-μm core diameter and a length of 300 m, approximately 800 propagation modes can be supported. Theoretically, this could reduce optical crosstalk noise by nearly 29-fold without requiring any active control electronics.
"This is not the end of the journey. We already know what needs to be improved: speed, data volume, phase encoding, and reconstruction automation. However, the underlying concept and the way we describe the phenomenon offer tremendous opportunities for the development of both new imaging devices and further innovations of the method itself," concludes Prof. Maciej Wojtkowski.
Maciej Wojtkowski (2026). Spatio-temporal optical coherence imaging and tomography for in vivo applications. Journal of Biomedical Optics.
DOI: https://doi.org/10.1117/1.JBO.31.11.113504
Author: Scientific Editor Marcin Powęska
