What is Spectral Domain OCT?

In SD-OCT, light from a broadband source is divided into two paths: one directed toward the sample and the other toward a reference mirror. The reflected light from these two paths interferes, producing a wavelength-dependent interference pattern (spectrum).

Reflections from specific depths correspond to distinct harmonic frequencies in k-space (2π/λ). Consequently, the depth profile can be reconstructed by applying an Inverse Fourier transform to the measured spectrum in k-space.

The depth profile of a sample obtained at a single scan point is referred to as an A-scan. A sequence of A-scans taken along a line across the sample forms a B-scan, which represents a cross-sectional image of the sample. By recording B-scans along a perpendicular dimension of the sample, a complete volumetric image—known as a C-scan—can be generated.

Spectral-domain optical coherence tomography (SD-OCT) offers significant advantages in high-resolution imaging, particularly in biomedical applications. It provides non-invasive, depth-resolved visualization of tissue structures, particularly in ophthalmology for diagnosing retinal diseases like glaucoma and age-related macular degeneration. SD-OCT delivers superior axial resolution due to its broad spectral bandwidth and allows volumetric imaging (C-scans), offering comprehensive 3D tissue visualization.

One of SD-OCT’s key advantages is its fast acquisition speed, capturing tens of thousands to hundreds of thousands of A-scans per second. A full 3D retinal scan can be completed in 1–5 seconds, significantly reducing motion artifacts and minimizing patient discomfort. In contrast, fluorescence tomography, which relies on molecular absorption and emission, requires several minutes for full volumetric imaging.

An SD-OCT system relies on a broadband light source, an interferometer, a scanner, a spectrometer and a processing unit.

The light source provides a range of wavelengths – typically in the infrared – with a bandwidth of 50 – 200 nm. A common type of light source is a super luminescent diode (SLD). Typical center wavelengths used are 840 nm, 880 nm, 1050 nm, 1310 nm, and 1550 nm.

The interferometer splits light from the source, directing it toward both the sample and a reference mirror. It then recombines the reflected light from these two paths, generating an interference pattern. The interferometer can be designed using either free-space optics or fiber optics.

The scanner is responsible for directing the light beam across the sample. This is typically achieved using galvanometric mirrors, MEMS (micro-electromechanical systems) scanners, or piezoelectric devices.

The spectrometer captures the interference patterns as a function of wavelength by using a high-dispersion diffraction grating and a line scan camera. The frame rate of the camera is typically in the kHz region to enable short measurement time.

In SD-OCT, system performance is governed by several interconnected parameters, including axial and lateral resolution, imaging depth, spectral bandwidth, and pixel count. Understanding these relationships helps optimize imaging quality and system design.

The axial resolution (Δz) in SD-OCT is determined by the coherence length of the source and is inversely related to the spectral bandwidth (Δλ) of the superluminescent diode (SLD) or other broadband light sources:

where:

• λ₀ is the central wavelength of the light source
• Δλ is the spectral bandwidth (FWHM),
• The factor 2ln2 / π accounts for the coherence function’s shape for a Gaussian profile.

Please, note that the above formula is true for an SLED with a Gaussian spectral profile. For a flat top profile Δz is about 20% larger. In any case, it can be seen that a broader spectral bandwidth improves axial resolution.

Lateral resolution (Δx) is primarily determined by the focusing optics and numerical aperture (NA) of the system:

Increasing the numerical aperture improves lateral resolution but reduces the depth of focus.

The imaging depth of an SD-OCT system depends on the penetration depth at the wavelength used. Furthermore, the maximum imaging depth (zmax) is influenced by the spectral bandwidth and number of pixels in the spectrometer:

where:

• N_pixels is the number of detector pixels in the spectrometer and Δλ _spectrometer is the bandwidth of the spectrometer.

In practice, the maximum number of pixels is typically limited by commercially available detectors.

Sensitivity roll-off refers to the gradual reduction in signal strength at greater imaging depths, caused by limitations in spectral resolution and fringe visibility. This effect leads to decreased contrast and signal quality for structures located further from the zero-delay position.

The axial sampling density depends on the number of pixels in the spectrometer and the spectral range:

The depth profile is reconstructed by applying the Fourier transform to the acquired spectral data, converting k-space information into spatial domain representations.

To accurately reconstruct the depth profile, the system must satisfy the Nyquist sampling criterion, which states that the spectral sampling interval must be small enough to resolve all spatial frequencies present in the signal. This means the number of detector pixels in the spectrometer (N_pixels) must be sufficiently high to avoid aliasing and ensure proper depth resolution:

where:

• Δk_min is the minimum spectral sampling interval required,
• z_max is the maximum imaging depth.

A higher pixel count in the spectrometer enables better spectral resolution, thereby extending the imaging depth while maintaining signal fidelity.

Optimizing an SD-OCT system requires balancing resolution, imaging depth, and sensitivity. A larger bandwidth enhances axial resolution but demands a higher pixel count in the spectrometer to maintain depth range. Similarly, a higher numerical aperture improves lateral resolution but reduces the imaging depth.

Spectral-domain optical coherence tomography (SD-OCT) is widely used in ophthalmology, providing high-resolution imaging for diagnosing and monitoring retinal diseases, such as glaucoma and macular degeneration. It is also valuable in dermatology for assessing skin layers non-invasively and in cardiology for visualizing vascular structures and plaques.

Beyond medicine, SD-OCT plays a role in material science and biological research, helping analyze microstructures with precision. Its ability to deliver real-time, depth-resolved imaging makes it a powerful tool across multiple fields.

Click here for a detailed exploration of SD-OCT applications.