Optical Coherence Tomography (OCT): How It Works
A single OCT scan captures a cross-sectional image of the retina with axial resolution as fine as 5–7 micrometers — roughly one-tenth the width of a human hair. That level of detail, achieved without touching the eye or injecting dye, has made OCT the most frequently performed imaging procedure in ophthalmology. More than 30 million OCT scans are obtained annually in the United States alone (National Eye Institute). Understanding how the technology produces those images helps clarify why it has become indispensable for diagnosing and managing conditions from glaucoma to diabetic retinopathy.
The Core Principle: Low-Coherence Interferometry
OCT borrows a concept from physics that dates to the 19th century — interferometry — and applies it to biological tissue using near-infrared light, typically in the 800–1,050 nanometer wavelength range. A broadband light source emits a beam that is split into two paths. One path, the sample arm, directs light into the eye. The other, the reference arm, bounces light off a mirror at a known distance.
When the two beams recombine, they create an interference pattern only when their optical path lengths match within the coherence length of the light source. Because that coherence length is extremely short (a few micrometers), the system can pinpoint exactly how deep within tissue each reflection originated. This is low-coherence interferometry — essentially using the light's own statistical properties as a ruler.
The result is a depth-resolved reflectivity profile called an A-scan (analogous to ultrasound A-mode). Thousands of adjacent A-scans are stitched together to form a B-scan: the familiar cross-sectional image that shows the layered architecture of the retina, from the nerve fiber layer at the surface down through the retinal pigment epithelium and into the choroid.
Time-Domain vs. Spectral-Domain: A Speed Revolution
The earliest clinical OCT systems, introduced commercially by Carl Zeiss Meditec in 1996, used time-domain (TD-OCT) technology. In TD-OCT, the reference mirror physically moves back and forth to scan different depths, one point at a time. The approach worked, but it was slow — about 400 A-scans per second — and motion artifacts were a constant problem.
Spectral-domain OCT (SD-OCT), which became widely available around 2006, eliminated the moving mirror entirely. Instead of measuring one depth at a time, SD-OCT captures all depths simultaneously by spreading the returning light across a spectrometer with a charge-coupled device (CCD) or CMOS sensor. A mathematical operation called a Fourier transform converts the spectral interference pattern into a full depth profile in one shot.
The speed gain is dramatic: SD-OCT systems routinely acquire 20,000–100,000 A-scans per second (American Academy of Ophthalmology). That 50- to 250-fold improvement over TD-OCT means denser scan patterns, fewer motion artifacts, and three-dimensional volumetric imaging of the macula and optic nerve head.
Swept-Source OCT: Going Deeper
Swept-source OCT (SS-OCT) represents a further refinement. Rather than a broadband light source paired with a spectrometer, SS-OCT uses a tunable laser that rapidly sweeps through a range of wavelengths — often centered around 1,050 nm instead of the 840 nm typical of SD-OCT. The longer wavelength penetrates deeper into tissue and scatters less in media opacities such as cataracts.
SS-OCT systems can exceed 200,000 A-scans per second, enabling wide-field imaging of both the retina and the choroid in a single acquisition. This has opened new clinical territory for studying conditions like central serous chorioretinopathy and pachychoroid spectrum diseases, where the pathology lies beneath the retinal pigment epithelium (Cleveland Clinic).
OCT Angiography: Motion as Contrast
One of the most significant extensions of OCT technology is OCT angiography (OCTA), which detects blood flow without injected dye. OCTA works by comparing consecutive B-scans taken at the same retinal location. Stationary tissue produces identical signals, while moving red blood cells create decorrelation — a measurable change between scans. That decorrelation signal maps the microvasculature.
OCTA can separately visualize the superficial capillary plexus, the deep capillary plexus, and the choriocapillaris — distinct vascular beds that are impossible to isolate cleanly on a fluorescein angiogram. The technique has proven particularly valuable for detecting diabetic macular ischemia and quantifying the foveal avascular zone, which averages approximately 0.25–0.35 mm² in healthy eyes (NEI research data).
What OCT Actually Shows the Clinician
A standard macular OCT scan displays roughly 10 distinct retinal layers. Ophthalmologists look for specific disruptions: subretinal fluid in age-related macular degeneration, cystoid spaces in diabetic macular edema, thinning of the retinal nerve fiber layer (RNFL) in glaucoma. Normative databases built into commercial platforms compare a patient's RNFL thickness — typically around 90–105 micrometers in healthy adults — against age-matched controls, flagging statistically significant thinning with color-coded probability maps (Bascom Palmer Eye Institute / University of Miami).
The scan itself takes roughly 3–5 seconds per eye. No pupil dilation is strictly required for most protocols, though dilation improves image quality. There is no ionizing radiation, no contact with the cornea, and no known biological risk from the near-infrared light levels used.
Limitations Worth Knowing
OCT is not without blind spots. Dense cataracts and vitreous hemorrhage can degrade signal quality significantly because the light cannot reach or return from the retina cleanly. OCTA, while dye-free, has a limited field of view compared to traditional fluorescein angiography — typically 3×3 mm to 12×12 mm — and cannot detect leakage, only flow. And like any imaging modality, OCT produces data that requires skilled interpretation; an artifact mistaken for pathology can lead to unnecessary treatment.
FAQ
Does an OCT scan hurt?
No. The procedure is entirely non-contact and non-invasive. The patient rests the chin on a support and fixates on a target while near-infrared light — invisible and imperceptible — scans the eye.
How often should OCT be performed for glaucoma monitoring?
Practice guidelines from the American Academy of Ophthalmology suggest OCT imaging at baseline and at intervals determined by disease severity, often every 6–12 months for stable glaucoma suspects and more frequently for progressive disease (AAO Preferred Practice Patterns).
Can OCT replace fluorescein angiography?
For certain indications, OCTA has reduced the need for dye-based angiography. However, fluorescein angiography remains essential when leakage assessment or ultra-wide-field vascular imaging is required, particularly in proliferative diabetic retinopathy and retinal vein occlusions.
Is OCT used outside ophthalmology?
Absolutely. Cardiology uses intravascular OCT to image coronary artery plaques, and dermatology applies it to evaluate skin lesions. The underlying physics is identical — only the probe design and wavelength change.
References
- National Eye Institute — Optical Coherence Tomography
- American Academy of Ophthalmology — What Is OCT?
- American Academy of Ophthalmology — Preferred Practice Patterns
- Cleveland Clinic — Optical Coherence Tomography
- National Eye Institute — Diabetic Retinopathy
- Bascom Palmer Eye Institute, University of Miami
The law belongs to the people. Georgia v. Public.Resource.Org, 590 U.S. (2020)