1. What is Optical Coherence Tomography (OCT)?

Definition and Core Concept

Optical Coherence Tomography (OCT) is a non-invasive, high-resolution, cross-sectional imaging modality that uses low-coherence interferometry to capture micrometer-scale structural information from biological tissues in real time. Unlike conventional optical microscopy, OCT achieves depth-resolved imaging by exploiting the coherence properties of broadband light — typically in the near-infrared spectrum — to discriminate backscattered photons originating from specific tissue depths. This process, known as coherence gating, allows the system to reject multiply scattered light and reconstruct precise depth profiles without physical sectioning of the tissue. Axial (depth) resolution typically ranges from 1 to 15 micrometers depending on the light source bandwidth and center wavelength, while lateral resolution is governed by the focusing optics of the sample arm. Because OCT operates entirely with light, it is inherently free of ionizing radiation and requires no exogenous contrast agents, making it exceptionally well-suited for in vivo clinical imaging and intraoperative guidance.

Historical Background

The foundational principles underlying OCT trace back to optical low-coherence reflectometry research conducted at Bell Laboratories in the 1970s, where scientists explored coherence-based methods for characterizing optical fiber components. Parallel advances in femtosecond laser technology and broadband light sources throughout the 1980s provided the enabling photonic infrastructure. The pivotal breakthrough came in 1991 when David Huang, James Fujimoto, and colleagues at MIT and Harvard Medical School published the landmark paper in Science demonstrating OCT imaging of the human retina and coronary artery ex vivo with ~15 μm resolution. This publication established OCT as a viable biomedical imaging technology. Clinical translation accelerated through the early 1990s, culminating in the first commercially available OCT system — the Zeiss OCT 1 — receiving regulatory clearance and entering ophthalmic practice in 1996. Subsequent decades witnessed rapid evolution from time-domain to Fourier-domain architectures, dramatically increasing imaging speed and sensitivity.

OCT vs Other Imaging Modalities

OCT occupies a unique niche in the biomedical imaging landscape, offering capabilities that complement rather than replace existing modalities. Compared to diagnostic ultrasound, OCT uses light rather than acoustic waves, yielding resolution 10–100 times superior (micrometers vs. hundreds of micrometers) but with reduced penetration depth (1–3 mm vs. several centimeters in tissue) due to strong optical scattering. Unlike CT scanners, OCT produces no ionizing radiation and poses no cumulative dose risk to patients, though it cannot image through bone or deep organs. Relative to MRI, OCT systems are substantially faster (real-time imaging at tens of frames per second), more compact, less expensive, and do not require magnetic shielding or superconducting infrastructure — though again at the cost of penetration depth. For practitioners working across radiological imaging domains, OCT serves as a complementary micro-scale tool for superficial tissue characterization where macroscale modalities lack sufficient spatial resolution.

2. Why is OCT Used?

Clinical Need

The driving clinical imperative for OCT adoption stems from the need to visualize tissue microarchitecture in vivo without the delays, costs, and risks associated with excisional biopsy or invasive diagnostic procedures. In ophthalmology — OCT’s most mature application domain — retinal diseases such as age-related macular degeneration (AMD), diabetic macular edema (DME), glaucoma, and epiretinal membranes present with subtle structural changes in the retinal layers that are clinically invisible using indirect ophthalmoscopy alone. OCT enables detection and quantification of these changes at the level of individual retinal sublayers with micron-scale precision, making it the standard of care in retinal practice worldwide. Beyond ophthalmology, there exists a compelling clinical need for real-time, high-resolution guidance in cardiovascular interventions, where intravascular OCT (IVOCT) characterizes coronary artery plaque morphology and stent deployment with far greater detail than intravascular ultrasound. Oncological and dermatological applications further leverage OCT as an optical biopsy tool for identifying dysplastic tissue without surgical excision.

Key Advantages

OCT delivers a combination of attributes that no single competing modality replicates: (1) Non-invasiveness — imaging is performed through natural body openings or transparent tissue windows without contact or incision; (2) High resolution — axial resolution of 1–15 μm enables visualization of cellular and subcellular tissue layers; (3) Real-time acquisition — modern Fourier-domain systems acquire 20,000 to 400,000 A-scans per second, enabling live cross-sectional imaging during clinical examination or surgical procedures; (4) No ionizing radiation or contrast agents — OCT is safe for repeated imaging in sensitive populations including neonates and pregnant patients; (5) Quantitative measurement — automated segmentation algorithms provide reproducible thickness maps of retinal layers, corneal stroma, and vascular wall structures, supporting longitudinal disease monitoring and treatment response assessment; (6) Compact and portable potential — fiber-optic-based architectures enable miniaturization into handheld, endoscopic, and catheter-based probe configurations.

Regulatory Status

OCT devices are regulated in the United States by the Food and Drug Administration (FDA) under 21 CFR Part 886 for ophthalmic devices, classified as Class II medical devices subject to 510(k) premarket notification. The first FDA clearance for an ophthalmic OCT system was granted in 1996 for the Zeiss OCT 1 system. Since then, hundreds of 510(k) clearances have been issued across ophthalmic, intravascular, and dermatological OCT platforms. Intravascular OCT systems are regulated under cardiovascular device classifications. In the European Union, OCT systems bear the CE mark under the Medical Device Regulation (MDR 2017/745), with conformity assessment conducted by notified bodies. Internationally, devices must comply with IEC 60601-1 (general safety and performance) and ISO 15004-2 (ophthalmic instrument optical radiation safety limits). The well-established regulatory pathway and substantial post-market clinical evidence base have accelerated institutional adoption and reimbursement coverage across major healthcare markets.

3. How Does OCT Work?

Physics of Low-Coherence Interferometry

OCT is fundamentally built upon the Michelson interferometer, adapted to exploit the short temporal coherence length of broadband light sources. In a standard Michelson configuration, light from a broadband source is split by a beam splitter into two paths: a reference arm directed toward a mirror at a known optical path length, and a sample arm directed toward the biological tissue under examination. Backscattered light from both arms recombines at the beam splitter and produces an interference pattern only when the optical path length difference between the two arms falls within the coherence length (l_c) of the source — typically a few micrometers. This phenomenon, termed coherence gating, provides the depth-discrimination mechanism. The axial (depth) resolution of an OCT system is directly determined by the coherence length and is approximated by the formula: Δz ≈ (2ln2/π) × (λ₀²/Δλ), where λ₀ is the center wavelength and Δλ is the full-width-at-half-maximum spectral bandwidth. A broader spectral bandwidth therefore yields finer axial resolution, independent of focusing optics. Lateral resolution, in contrast, is governed by the numerical aperture of the objective lens in the sample arm, analogous to conventional optical microscopy.

Imaging Process Step by Step

The OCT imaging sequence begins with the emission of broadband near-infrared light from the source, which is coupled into a single-mode optical fiber and directed to the fiber-optic beam splitter (fiber coupler). The reference arm light travels to a stationary retroreflecting mirror, while the sample arm light is collimated, directed through galvanometer-mounted scanning mirrors, and focused onto the tissue surface via an objective lens. Photons penetrating the tissue are backscattered from refractive index discontinuities at tissue layer interfaces — corresponding to changes in tissue microstructure. These backscattered photons return through the sample arm and recombine with reference arm light at the coupler. The resulting interference signal encodes depth-resolved reflectivity information. In time-domain OCT (TD-OCT), the reference mirror is mechanically translated to vary the path length and sequentially sample each depth. In Fourier-domain OCT (FD-OCT), all depths are acquired simultaneously by analyzing the spectral interference pattern, dramatically improving speed and sensitivity.

Signal Acquisition and Reconstruction

A single depth profile (A-scan) is constructed by detecting the interference fringe pattern and applying a Fourier transform to convert spectral domain data into a reflectivity-versus-depth plot. Sequential lateral scanning with the galvanometer mirrors across one axis produces a two-dimensional cross-sectional image called a B-scan (analogous to an ultrasound B-mode image). Raster scanning across two lateral axes yields a three-dimensional volumetric dataset from which arbitrary en face (C-scan) slices can be reconstructed. Raw interferometric data requires substantial digital signal processing: dispersion compensation, spectral resampling from wavelength to wavenumber space (for SD-OCT), fast Fourier transformation (FFT), logarithmic compression, and speckle noise reduction through frame averaging or compounding. Modern clinical OCT systems perform these operations in real time using GPU-accelerated processing pipelines, delivering live B-scan video at 25–100 frames per second for immediate clinical interpretation.

4. Main Components of an OCT System

Light Source

The light source is the most critical determinant of OCT system performance, defining axial resolution, penetration depth, and imaging speed. Superluminescent diodes (SLDs) are the workhorse source for spectral-domain OCT systems targeting the retina, operating at center wavelengths of 800–870 nm with bandwidths of 40–100 nm, yielding axial resolutions of 5–7 μm in tissue. For deeper tissue applications — including choroidal imaging, anterior segment OCT, and intravascular OCT — longer wavelengths of 1050–1310 nm are preferred due to reduced scattering in melanin-containing and fibrous tissues. Swept-source OCT (SS-OCT) systems employ tunable laser sources (MEMS-VCSEL or polygon-mirror-based) that rapidly sweep through wavelengths at rates of 100 kHz to 400 kHz, enabling ultra-high-speed volumetric acquisition. Supercontinuum lasers and titanium:sapphire femtosecond lasers are used in research-grade ultrahigh-resolution systems achieving sub-micron axial resolution, though their size and cost currently preclude routine clinical deployment.

Interferometer and Beam Splitter

Clinical OCT systems universally employ fiber-optic Michelson interferometers rather than free-space bulk-optic designs, capitalizing on the mechanical stability, compact form factor, and alignment-free assembly of single-mode fiber components. The 50:50 fiber coupler functions as the beam splitter, dividing source light between the reference and sample arms and recombining returning light for detection. The reference arm contains a fiber-coupled retroreflector or mirror mounted on a motorized translation stage (in TD-OCT) or at a fixed path length (in FD-OCT), along with dispersion-compensating optical elements matched to the sample arm. A polarization controller in the reference arm optimizes fringe contrast by matching the polarization state of reference and sample light. In balanced detection configurations used in SS-OCT, a 2×2 coupler outputs two complementary interference signals to a dual-balanced photodetector pair, effectively doubling the signal and suppressing common-mode intensity noise from the source.

Detection and Signal Processing

Detection architecture differs fundamentally between OCT subtypes. Spectral-domain OCT (SD-OCT) disperses the combined interference light using a diffraction grating onto a high-speed line-scan CCD or CMOS camera (silicon arrays for 800 nm; InGaAs arrays for 1050 nm), acquiring the full spectral interferogram simultaneously. Camera line rates of 70,000–250,000 lines per second determine maximum A-scan acquisition speed. Swept-source OCT (SS-OCT) instead uses a single or dual-balanced photodetector pair with high bandwidth (>1 GHz), sequentially detecting interference as the laser wavelength sweeps in time. Analog-to-digital converters (ADCs) digitize the photodetector output at rates up to 1.6 GS/s. Downstream signal processing — implemented on FPGA and GPU hardware — performs real-time FFT, dispersion correction, and image rendering. Deep learning-based denoising and segmentation algorithms are increasingly integrated into commercial platforms, automating retinal layer thickness quantification and pathology detection.

Sample Arm and Scanning Optics

The sample arm interfaces the fiber-optic system with the biological tissue under examination and determines lateral resolution and field of view. Light exiting the fiber is collimated by an achromatic lens, directed through a pair of orthogonally oriented galvanometer scanning mirrors (X-Y scanners) for lateral beam steering, and focused onto the tissue by an objective lens optimized for the operating wavelength. For ophthalmic OCT, the patient’s own cornea and lens serve as part of the focusing optical system, requiring the device to be calibrated for the optical power of the eye. Slit-lamp delivery systems, handheld probes, and surgical microscope-integrated configurations exist for different clinical contexts. For intravascular and endoscopic applications, the sample arm is miniaturized into a rotary catheter (1.8–3.5 French) containing a gradient-index (GRIN) lens and a single-mode fiber with a distal microprism, enabling 360° rotational scanning of luminal structures at up to 100 frames per second during catheter pullback.

5. Types and Variants of OCT

Time-Domain vs Fourier-Domain OCT

The evolution from time-domain to Fourier-domain OCT represents the single most transformative advance in OCT engineering history, delivering improvements of 50–100× in imaging speed and 20–30 dB in sensitivity simultaneously. In TD-OCT, depth scanning is achieved by mechanically translating the reference mirror, limiting acquisition to approximately 400 A-scans per second — insufficient for motion-artifact-free volumetric imaging of the living eye. Fourier-domain approaches (both SD-OCT and SS-OCT) encode all depth information in the spectral interference pattern acquired at a single moment, eliminating mechanical scanning of the reference arm and enabling acquisition rates of 20,000 to 400,000 A-scans per second. The sensitivity advantage of FD-OCT arises from Fellgett’s multiplex advantage: all depths contribute simultaneously to the detected signal, whereas TD-OCT detects only one depth at a time. This enables FD-OCT to operate with lower source power while achieving superior signal-to-noise ratio — a critical safety consideration in ophthalmic imaging governed by ANSI Z136.1 maximum permissible exposure limits.

Specialized OCT Modalities

Beyond structural imaging, functional OCT extensions expand diagnostic capability into hemodynamics, biomechanics, and molecular contrast. OCT Angiography (OCTA) detects moving red blood cells by analyzing inter-scan decorrelation of speckle patterns across repeated B-scans at the same location, generating depth-resolved microvascular flow maps of the retina and choroid without fluorescent dye injection — transforming the assessment of diabetic retinopathy and neovascular AMD. Polarization-Sensitive OCT (PS-OCT) measures the polarization state change of backscattered light to map birefringent tissue structures including the retinal nerve fiber layer, corneal collagen, and myelin, providing contrast unavailable in conventional intensity-based OCT. Doppler OCT extracts quantitative blood flow velocity from phase-resolved interference signals, enabling retinal vessel flowmetry. Elastography OCT combines OCT with mechanical perturbation to map tissue stiffness at micrometer scales, with applications in early cancer detection and corneal biomechanics. The following table summarizes the principal OCT system types and their performance characteristics: