Speckle variance optical coherence tomography

Imaging system

SV-OCT can be done with spectral domain OCT (SD-OCT) and swept source OCT (SS-OCT).^[5] SD-OCT and SS-OCT are both methods of Fourier domain OCT (FD-OCT), which has significantly faster image acquisition speed compared to time domain OCT. In general, OCT measures the echo time delay and intensity of reflected and backscattered light. A broad-bandwidth laser or superluminescent diode low-coherence light source travels to a beam splitter, which sends half of the light to the reference arm, which is at a known location, and half of the light to the sample, where it scatters and reflects off tissue. Light from the reference and sample arms recombine at the beam splitter, forming an interference pattern that is sensed by a photodetector. In SD-OCT, the interference pattern is split into its frequency components by a grating and are simultaneously detected by a charge-coupled device (CCD). Each frequency corresponds to a certain depth within the tissue.^[6] In SS-OCT, a tunable swept laser source is used.^[7][8]

Algorithm

The intensity or speckle of an OCT signal is the random interference pattern produced by backscattered light from a random medium. OCT captures cross-sectional images, known as B-scans. In SV-OCT, multiple B-scans are captured at the same location, creating a 3D data set, with time as the third dimension. The pixel-wise variance is computed between consecutive B-scan frames. A speckle variance image, ${\displaystyle I_{SV}}$ is calculated as

${\displaystyle I_{SV}={\frac {1}{N}}\sum _{i=1}^{N}\left(I_{ijk}-{\frac {1}{N}}\sum _{i=1}^{N}I_{ijk}\right)^{2}\qquad \quad (1)}$

where ${\displaystyle {N}}$ is the number of B-scans obtained at a single location and ${\displaystyle I_{ijk}}$ is the intensity of a pixel with image coordinates ${\displaystyle (j,k)}$ in the B-scan indexed by ${\displaystyle i}$.

The speckle pattern of OCT images are affected by the motion of scattering particles in the target medium. The interference pattern produced by the signal of backscattered light through a medium depends on the movement of these particles. Therefore, the speckling pattern encapsulates information regarding the spatial and temporal motion of scattering particles in a random scattering medium.^[9] SV-OCT uses the inter-frame variance of image intensities to compute blood flow. Areas that have high flow will have higher motion of scattering particles and this information is encoded in the speckle pattern.

SV-OCT has advantages for microvasculature imaging due to its high sensitivity and independence to the Doppler angle. In addition, it has low computational complexity and requires relatively low data storage compared to PV-OCT.^[10] However, SV-OCT it is susceptible to bulk tissue motion and multiple scattering induced artifacts.^[11]