Optical coherence tomography

Based on Wikipedia: Optical coherence tomography

Seeing Inside the Body with Light

Imagine being able to peer beneath the surface of living tissue—watching a retina in real time, examining the walls of a coronary artery from the inside, or mapping the layers of skin without ever making a cut. This isn't science fiction. It's happening millions of times a year in clinics around the world, thanks to a technology that essentially turns light into a kind of sonar.

The technology is called optical coherence tomography, or OCT. And if you've never heard of it, that's somewhat remarkable, because by 2016 it was being used in more than 30 million imaging procedures annually worldwide. It has transformed how doctors diagnose and monitor the three leading causes of blindness—macular degeneration, diabetic retinopathy, and glaucoma—preventing vision loss in countless patients.

So how does shining light into tissue let you see what's hidden beneath the surface?

The Core Idea: Optical Ultrasound

The simplest way to understand OCT is to think of it as "optical ultrasound." In traditional ultrasound imaging, doctors send sound waves into your body and listen for echoes. Different tissues reflect sound differently, and by measuring how long the echoes take to return, the machine can build up a picture of what's inside you.

OCT does essentially the same thing, but with light instead of sound.

This matters because light waves are much shorter than sound waves. Sound waves used in medical ultrasound are typically a few hundred micrometers long, which limits the detail you can see. Light waves are roughly a thousand times shorter—around a micrometer or less. This means OCT can resolve structures that are just a few micrometers across, approaching the scale of individual cells.

But there's a catch. When you shine light into tissue, it scatters in all directions. Imagine shining a flashlight into fog—you see a diffuse glow, not a clear beam. If you simply tried to photograph tissue illuminated from within, you'd get a blurry mess. The light bouncing around chaotically would overwhelm any useful signal.

OCT solves this problem through a clever trick borrowed from physics: interferometry.

Interferometry: The Art of Measuring Light's Travel Time

Interferometry is a technique for measuring distances using the wave nature of light. It relies on a phenomenon called interference—when two light waves meet, they can either reinforce each other (constructive interference, making brighter light) or cancel each other out (destructive interference, making dimmer light or darkness).

Here's how OCT uses this principle. The system splits a beam of light into two paths. One path—the sample arm—directs light into the tissue being examined. The other path—the reference arm—bounces light off a mirror at a known distance. Both beams then recombine at a detector.

When light reflected from inside the tissue travels exactly the same distance as light from the reference mirror, the two beams interfere with each other, creating a detectable signal. When the distances don't match, the interference pattern vanishes.

By systematically changing the position of the reference mirror, you can probe different depths within the tissue. Each depth that produces an interference signal tells you something is there—a boundary between tissues, a blood vessel, a structural feature. The strength of the signal tells you how reflective that feature is.

This technique is called low-coherence interferometry, and it gives OCT its remarkable depth discrimination. Only light that has traveled a very specific distance contributes to the image. All that scattered light bouncing around chaotically? It travels different distances and doesn't produce interference, so it's automatically filtered out.

What Makes the Light "Low Coherence"?

The word "coherence" in optical coherence tomography refers to how orderly the light waves are. A laser produces highly coherent light—all the waves march in lockstep over long distances. This is why lasers can create interference patterns across meters or even kilometers.

OCT deliberately uses light with low coherence—light that only stays "in step" with itself over very short distances, typically just a few micrometers. This is achieved by using light sources that emit many different wavelengths simultaneously, rather than the single pure wavelength of a laser.

Superluminescent diodes are a common choice. These devices produce bright light spread across a range of wavelengths, giving them the low coherence needed for OCT. Femtosecond lasers—lasers that produce extremely short pulses lasting only quadrillionths of a second—also work, because their ultrabrief pulses inherently contain many wavelengths.

The lower the coherence, the finer the depth resolution. If your light stays coherent only over a distance of two micrometers, you can distinguish features that are just two micrometers apart in depth. This is what gives OCT its subcellular precision.

Building an Image: From A-Scans to Three Dimensions

A single measurement with OCT produces what's called an A-scan—a one-dimensional profile of reflectivity versus depth, like a single vertical line through the tissue. This tells you what structures exist at different depths along that line.

To create a two-dimensional cross-sectional image—called a B-scan—the system scans the light beam sideways across the tissue, collecting A-scans at many adjacent positions. Combine all these vertical lines, and you get a slice through the tissue, similar to what you'd see if you cut the tissue open and looked at it edge-on.

Take many parallel B-scans, and you can build a complete three-dimensional volume. Modern OCT systems can acquire these volumes remarkably quickly—hundreds of thousands of depth scans per second, allowing full 3D imaging in a fraction of a second.

The Revolution in Speed

When OCT was first demonstrated in 1991 by a team from the Massachusetts Institute of Technology and Harvard Medical School, the original system could manage only about 0.8 depth scans per second. This was far too slow for practical clinical use on living patients, who move and blink.

The breakthrough came from a shift in how the measurements are performed. The original "time-domain" OCT systems physically moved the reference mirror to scan through different depths—a mechanical process with inherent speed limits.

Fourier-domain OCT, which became clinically available around 2006, eliminated this mechanical scanning. Instead of moving a mirror, these systems analyze all depths simultaneously by looking at how different wavelengths of light interfere. Using mathematical techniques based on the Fourier transform—a way of decomposing signals into their constituent frequencies—the system can extract depth information from a single measurement.

The speed improvement was dramatic. Fourier-domain systems can acquire tens of thousands to hundreds of thousands of depth scans per second, more than enough for real-time imaging of living tissue. Laboratory prototypes now operate at multiple millions of scans per second.

Over three decades, commercial clinical OCT systems have increased in speed more than a thousand-fold, doubling their performance roughly every three years. This pace rivals Moore's Law—the famous observation about computer chip performance doubling every two years.

What OCT Can and Cannot See

Every imaging technology has its strengths and limitations. OCT excels at high-resolution imaging of shallow structures.

The resolution is remarkable: current systems achieve about 10 micrometers axially (in depth) and around 20 micrometers laterally. For context, a typical human cell is about 10 to 30 micrometers across, so OCT can resolve individual cell layers in many tissues.

The penetration depth, however, is limited. Light scattering in biological tissue means that useful signals can only be obtained from the top one to two millimeters. Below that depth, too little light survives without scattering to be detected.

This contrasts sharply with other imaging methods. Ultrasound and magnetic resonance imaging, or MRI, can see through the entire body, but their resolution is typically a fraction of a millimeter at best—a hundred times coarser than OCT. Confocal microscopy can achieve even finer resolution than OCT, down to sub-micrometer levels, but it can only see about 100 micrometers deep.

OCT occupies a unique niche: deeper than optical microscopy, sharper than whole-body imaging. This makes it ideally suited for tissues that are naturally accessible to light—especially the eye.

The Eye: OCT's Natural Home

The eye is almost perfectly designed for OCT imaging. Light can enter through the pupil and reach the retina at the back of the eye without obstruction. The cornea and lens are transparent. The retina itself is a thin, layered structure less than half a millimeter thick—well within OCT's penetration range.

Before OCT, examining the retina in fine detail required invasive procedures or was simply impossible in a living patient. Ophthalmologists could see the retinal surface through an ophthalmoscope, but the internal layers remained hidden.

OCT changed everything. For the first time, doctors could see cross-sections through the living retina, distinguishing its ten distinct layers. They could measure retinal thickness with micrometer precision, detect subtle swelling or thinning, identify fluid accumulation, and spot early damage from disease.

For macular degeneration—the leading cause of blindness in developed countries—OCT reveals whether abnormal blood vessels are leaking fluid beneath the retina, guiding decisions about treatment. For diabetic retinopathy, it shows thickening of the macula that threatens central vision. For glaucoma, it measures thinning of the nerve fiber layer, often detecting damage before patients notice any vision loss.

The impact has been so profound that October has been informally designated "OCT appreciation month" since 2012, with medical education events and journal clubs celebrating the technology's contributions to eye care.

Beyond the Eye: Seeing Inside Arteries

While ophthalmology remains OCT's largest application, the technology has found a second major home inside the cardiovascular system.

Intravascular OCT involves threading a tiny catheter through blood vessels until it reaches the coronary arteries—the vessels that supply blood to the heart muscle. At the tip of this catheter sits a miniature OCT system, smaller than a grain of rice, that can image the artery walls from the inside.

Why would cardiologists want to see artery walls at micrometer resolution? The answer lies in how heart attacks happen.

Most heart attacks aren't caused by arteries that gradually narrow until they close completely. Instead, they result from the rupture of vulnerable plaques—fatty deposits in the artery wall covered by a thin fibrous cap. When these caps break open, blood clots form rapidly, blocking blood flow suddenly and catastrophically.

Traditional imaging methods like angiography can show how narrow an artery has become, but they can't reveal the composition of the plaque or the thickness of its protective cap. OCT can. It can distinguish between stable plaques with thick fibrous caps and dangerous plaques with thin caps that might rupture. This information helps cardiologists decide how aggressively to treat patients and where to place stents.

The connection to cardiology is particularly timely. There's growing recognition in the field that focusing solely on "blockages"—obstructive narrowings that limit blood flow—may miss the bigger picture. Many heart attacks occur in arteries that weren't severely narrowed beforehand. The shift toward understanding atheroma (the fatty deposits themselves) and inflammation in artery walls represents a fundamental change in how cardiologists think about coronary disease.

Expanding Frontiers

OCT continues to find new applications beyond eyes and hearts.

In dermatology, it can image skin layers without biopsy, potentially allowing early detection of melanoma and other skin cancers by revealing abnormal tissue architecture. Gastroenterologists are exploring its use in the esophagus and other parts of the digestive tract, where it might detect early cancer or inflammatory changes. Neurologists and neurosurgeons are investigating its potential for imaging nerves and brain tissue. Dentists use it to examine tooth structure and detect decay.

One surprising application is in art conservation. Paintings are layered structures—varnish over paint over underdrawing over primer over canvas. OCT can reveal these hidden layers nondestructively, helping conservators understand an artwork's history and plan restoration efforts. The same technology that peers beneath your retina can peer beneath a Rembrandt's brushstrokes.

New Capabilities: Angiography Without Dye

The speed of modern OCT systems has enabled entirely new types of imaging that weren't possible with slower technology.

OCT angiography is one example. By acquiring multiple images of the same location in rapid succession, the system can detect subtle changes caused by flowing blood. Red blood cells moving through capillaries create tiny fluctuations in the optical signal. By analyzing these fluctuations, OCT can map blood vessel networks without injecting any dye into the patient.

Traditional retinal angiography requires injecting fluorescent dye into a vein, waiting for it to circulate to the eye, and then photographing its glow as it flows through retinal vessels. The dye occasionally causes allergic reactions, and the procedure takes time. OCT angiography achieves similar results in seconds with no injection at all.

Other novel capabilities include elastography—measuring the mechanical stiffness of tissue by how it deforms—and optoretinography, which can detect the optical changes that occur when photoreceptor cells in the retina respond to light. This latter technique is remarkable: it essentially allows researchers to watch individual cells seeing.

The Inventors and Their Recognition

The story of OCT's invention illustrates how breakthrough technologies often emerge from the convergence of multiple fields.

David Huang, then a graduate student, worked in James Fujimoto's laboratory at MIT. The Fujimoto lab had expertise in femtosecond lasers and their applications. Eric Swanson at MIT Lincoln Laboratory brought knowledge of fiber optics and signal processing. Collaborators at Harvard Medical School provided the clinical perspective and access to biological samples.

The 1991 paper in Science that introduced the term "optical coherence tomography" and demonstrated the first images was not created in isolation. Researchers worldwide had been exploring interferometry with low-coherence light throughout the 1980s. The potential for imaging had been proposed, and measurements of retinal thickness had been demonstrated. But the MIT-Harvard team brought these threads together into a practical imaging system and gave it a name that stuck.

For their contributions, Fujimoto, Huang, and Swanson received the 2023 Lasker-DeBakey Clinical Medical Research Award—often considered America's most prestigious biomedical research prize—as well as the National Medal of Technology and Innovation. It's a rare example of a medical imaging technology being recognized at this level, reflecting OCT's extraordinary impact on patient care.

The Physics in Plain Terms

For readers who want to understand the physics more deeply, here's a slightly more detailed explanation of how OCT extracts depth information.

When two light waves overlap, they interfere. If their peaks align, they add up to make a brighter combined wave—constructive interference. If the peak of one aligns with the trough of another, they cancel out—destructive interference. The pattern of bright and dark regions that results is called an interference pattern.

With highly coherent light (like a laser), this interference can occur even when the two beams have traveled very different distances. But with low-coherence light, interference only happens when the two paths are almost exactly equal—within a distance called the coherence length.

In OCT, one beam bounces off a reference mirror at a known distance. The other beam enters the tissue and reflects from various depths. Only the reflections that have traveled a distance matching the reference arm will produce interference. Everything else—all the multiply scattered light bouncing around chaotically—won't interfere and simply adds a small background noise.

By varying the reference distance (in time-domain OCT) or by analyzing how different wavelengths interfere differently (in Fourier-domain OCT), the system can determine what structures exist at each depth and how reflective they are.

The axial resolution depends on the coherence length, which in turn depends on how broad the spectrum of the light source is. Broader spectrum means shorter coherence length means finer resolution. The lateral resolution depends on how tightly the light beam can be focused, which is governed by the optics—entirely separate from the coherence properties.

This decoupling of axial and lateral resolution is one of OCT's distinctive features. Most imaging systems have the same resolution in all directions. OCT can have very fine axial resolution (from low coherence) while having somewhat coarser lateral resolution (from the focusing optics), or vice versa.

Safe and Gentle

OCT uses near-infrared or visible light at power levels that are completely safe for tissues. There's no ionizing radiation like X-rays or CT scans. No magnetic fields like MRI. No injected contrast agents in standard OCT imaging.

The examination is entirely non-contact for eye imaging—you simply look at a point of light while the machine scans your retina. There's no pressure on the eye, no drops required (though pupils are sometimes dilated for a wider view). Patients typically find it completely comfortable.

For intravascular imaging, the procedure is invasive since it requires threading a catheter through blood vessels, but this is part of the cardiac catheterization procedure that would be performed anyway for other diagnostic or therapeutic purposes. The OCT component adds no additional risk.

Looking Forward

The thousand-fold improvement in OCT speed over three decades raises an obvious question: what comes next?

Researchers are developing parallel acquisition approaches—line-field and full-field systems that capture multiple points simultaneously rather than scanning point by point. These could enable another leap in speed, continuing the Moore's Law-like trajectory.

Faster systems enable new applications. Real-time three-dimensional imaging during surgery. Detection of rapid physiological changes. Integration with artificial intelligence to automatically identify disease features.

The technology that started as a laboratory demonstration in 1991 has become one of the most widely used medical imaging technologies in the world. Thirty million procedures a year—and growing. For a technique most people have never heard of, that's a remarkable success story, written one photon at a time.