OCT is a noninvasive imaging method that uses low-coherence light and interferometry to reconstruct retinal structure in cross-section. Its power comes from repeatable, micron-scale structural information—not from producing a conventional photograph.
- OCT is a structural imaging technology: it maps optical reflectivity through depth rather than simply photographing the retinal surface.
- A single B-scan is only one slice of a larger dataset; modern OCT exams often contain dense volumes that can be segmented, measured and followed over time.
- Thickness maps and automated layer boundaries are useful, but they are software outputs that still require quality control and clinical interpretation.
OCT in one sentence
Optical coherence tomography, or OCT, is a light-based method for building cross-sectional images of tissue. In the retina, it converts differences in optical reflectivity into a depth-resolved view of layers that are otherwise superimposed when seen in a standard fundus photograph. The original 1991 OCT paper described low-coherence interferometry as a way to recover echoes of light from internal microstructures, analogous in concept—not in physics—to how ultrasound uses returning sound. That distinction matters: OCT is not an X-ray, does not use ionizing radiation, and is not simply a high-resolution camera. It is a measurement system that reconstructs where reflected light originated along depth.
From A-scans to B-scans to 3D volumes
The basic OCT building block is an A-scan, a one-dimensional reflectivity profile through depth. Place many A-scans beside one another and they form a B-scan: the familiar cross-sectional image in which the retina appears as layered bands. Repeat that process across an area and the system builds a three-dimensional volume. Modern retinal OCT therefore contains more information than the single screenshot most people recognize. Clinicians can scroll through adjacent slices, inspect en-face projections, compare registered follow-up scans and derive thickness or volume maps. When an AI model claims to read an 'OCT image,' an important follow-up question is whether it saw one selected B-scan or the full volume.
What the layers actually represent
OCT brightness reflects optical scattering, not a direct color code for anatomy. Different retinal interfaces and tissues return different amounts of light, producing alternating hyperreflective and hyporeflective bands. In a healthy macula, the foveal contour and outer-retinal bands create a recognizable pattern. Disease can alter those patterns through fluid, tissue loss, elevation, traction or deposits. But the image is an indirect representation: shadowing, signal loss and segmentation errors can change what appears on screen. This is why a structurally plausible OCT image is not automatically an accurate measurement.
Why OCT became indispensable in retina
OCT made subtle structural change measurable at the point of care. It can show intraretinal cystic spaces, subretinal fluid, pigment-epithelium detachments, macular holes, epiretinal membranes and thinning or disruption of retinal layers. Just as importantly, it can be repeated noninvasively, allowing clinicians and researchers to compare anatomy across visits. The clinical value is not that OCT replaces examination or every other modality. It is that it adds a precise structural dimension that photographs, angiography and visual acuity alone cannot provide.
Thickness maps are measurements, not universal constants
OCT systems often summarize anatomy with retinal thickness maps or automated layer measurements. Those numbers look objective, but they depend on device hardware, scan protocol and segmentation software. Comparative work has shown that central retinal thickness can differ significantly between OCT platforms even when agreement is otherwise high. Choroidal measurements can also vary between spectral-domain and swept-source systems. That means a change in device can create a change in number without an equivalent biological change. Longitudinal follow-up works best when acquisition and analysis are consistent.
Spectral-domain, swept-source and the direction of travel
Most contemporary retinal OCT belongs to spectral-domain or swept-source families. Both use Fourier-domain detection, but they differ in light source and acquisition architecture. Swept-source systems typically operate at longer wavelengths and can support faster scanning and deeper penetration into structures such as the choroid, while spectral-domain systems remain widely used and clinically capable. The useful question is not which acronym is 'better' in the abstract. It is whether a system's scan speed, wavelength, field of view, segmentation and repeatability match the clinical or research task.
What OCT cannot tell you by itself
OCT shows structure extremely well, but structure is not the same as blood-flow dynamics, dye leakage, cellular function or a complete diagnosis. Two diseases can produce similar structural patterns. A scan can also look quiet while another modality reveals vascular activity, or appear abnormal because segmentation failed. The safest mental model is complementary: OCT answers a powerful set of structural questions, while OCTA, fluorescein angiography, fundus autofluorescence, color imaging, examination and functional tests answer others.
What is next
OCT is becoming wider, faster and more computational. Dense volumetric analysis, swept-source widefield acquisition, home monitoring, automated fluid quantification and foundation-model approaches are pushing the modality beyond a single clinic snapshot. But every advance increases the importance of provenance and quality control: which device acquired the scan, what algorithm processed it, what part of the volume was analyzed and whether the result was externally validated. The future of OCT is not simply sharper pictures. It is better measurement—paired with better interpretation.
What the colors mean—and what they do not
Many OCT displays use false color to emphasize reflectivity or thickness, but the palette is a visualization choice rather than tissue color. A red region on a thickness map may mean 'thicker than a reference range,' not inflammation; a dark cavity on a grayscale B-scan may represent fluid because very little light returns from that space. Interpretation depends on knowing whether the display is structural reflectivity, a deviation map, segmentation overlay or en-face projection. The same anatomy can look very different when the display mode changes.
Repeatability is where OCT becomes a measurement tool
A powerful OCT workflow is not a single scan but a reproducible series. Eye tracking and registration can return to approximately the same retinal location on later visits, while software can compare thickness or lesion volumes over time. Repeatability is never perfect: fixation, scan density, segmentation and device changes matter. For research, this is why studies report reproducibility and test-retest variation. For readers, it is a reminder that a small numerical change is meaningful only when it exceeds expected measurement noise.
This guide explains OCT principles and interpretation at a publication level; it is not a substitute for device-specific training or clinical interpretation of an individual scan.
Sources & original records
We prioritize primary records, clinical-trial registries, peer-reviewed literature and authoritative institutions. Manufacturer material is labeled when used to describe a product or company position.
- Optical coherence tomographyScience / PubMed · 1991 · Foundational primary study · PMID 1957169 · DOI 10.1126/science.1957169
- Comparison of retinal thicknesses measured using swept-source and spectral-domain optical coherence tomography devicesPubMed · Comparative cohort study · PMID 25707041
- Comparison of choroidal thickness measurements between spectral-domain OCT and swept-source OCT in normal and diseased eyesPubMed · 2016 · Prospective comparative study · PMID 27881909 · DOI 10.2147/OPTH.S117022