Doppler optical coherence tomography

Abstract

Optical Coherence Tomography (OCT) has revolutionized ophthalmology. Since its introduction in the early 1990s it has continuously improved in terms of speed, resolution and sensitivity. The technique has also seen a variety of extensions aiming to assess functional aspects of the tissue in addition to morphology. One of these approaches is Doppler OCT (DOCT), which aims to visualize and quantify blood flow. Such extensions were already implemented in time domain systems, but have gained importance with the introduction of Fourier domain OCT. Nowadays phase-sensitive detection techniques are most widely used to extract blood velocity and blood flow from tissues. A common problem with the technique is that the Doppler angle is not known and several approaches have been realized to obtain absolute velocity and flow data from the retina. Additional studies are required to elucidate which of these techniques is most promising. In the recent years, however, several groups have shown that data can be obtained with high validity and reproducibility. In addition, several groups have published values for total retinal blood flow. Another promising application relates to non-invasive angiography. As compared to standard techniques such as fluorescein and indocyanine-green angiography the technique offers two major advantages: no dye is required and depth resolution is required is provided. As such Doppler OCT has the potential to improve our abilities to diagnose and monitor ocular vascular diseases.

Keywords: Angiography; Blood flow; Doppler effect; Optical Coherence Tomography; Perfusion; Retinal vasculature.

Fig. 1

Doppler OCT measures the axial velocity component v_‖. The yellow line represents the OCT beam.

Fig. 2

Upper panel: Measured time domain OCT A-scan of two static interface. middle panel: Time domain OCT signal of static interfaces with flow in-between causing a Doppler shift of the heterodyne carrier frequency. Sliding a time window (gray area) along the A-scan and calculating the centroid frequency allows extracting the local Doppler shift f_D (lower panel).

Fig. 3

(lhs) Histogram of phase difference plot of static tissue. (rhs) Histogram of Doppler shifted phase differences. If the distribution comes close to the limits of ±π the distribution will be wrapped by 2π causing and phase difference averaging leads to incorrect velocity values.

Fig. 4

Histogram (lhs) of the region confined by the black bars in phase difference tomogram (rhs). The region does not show any flow signal and can be used to quantify the phase difference fluctuations and thus the velocity sensitivity by the full-width-at-half-maximum of the histogram plot.

Fig. 5

Angle bandwidth for different absolute velocities and a given velocity bandwidth v_max − v_min (lhs). (rhs) schematic visualization of flow in dependence of flow velocities v_s and angle bandwidths.

Fig. 6

Geometrical situation at the posterior pole of the eye. K₁ and K₂ are the probe beam wave vectors, v is the velocity vector. α₁ and α₂ are the Doppler angles of the two laser beams and Δα is the separation angle between the two probe beams. γ is the angle between $\overrightarrow{v}$ and the plane perpendicular to the detection plane; and δ is the change in illumination angle due to eye movement (reproduced with permission from Werkmeister et al., 2012b).

Fig. 7

Phase Doppler tomograms of two orthogonal polarization channels corresponding to illumination directions along the indicated blue line. The fundus projection of a 3D data set allows extracting the angle β.

Fig. 8

Left panel: Fundus image and OCT en face image of a bifurcation measured with bidirectional Doppler OCT. The blue lines indicate the measurement locations. Right panel: Correlation between blood flow in trunk vessel and sum of blood flow in daughter vessels (R = 0.95; p < 0.001). (reproduced with permission from Werkmeister et al., 2012a).

Fig. 9

Blood flow versus blood vessel diameter on a log–log scale. Solid line: best fit result of linear regression (R = 0.72, p < 0.001). Dotted lines: 95% confidence interval. (reproduced with permission from Werkmeister et al., 2012a).

Fig. 10

Correlation analysis for velocity measurements. Correlation between velocity and blood flow as assessed with LDV and DOCT, respectively, during baseline conditions and during hyperoxia. Dotted lines: 95% confidence interval. (reproduced with permission from Werkmeister et al., 2012b).

Fig. 11

Calculation of angle ε. Angle ε as calculated from measurements using OCT (left panel) and LDV (right panel) from the two channels. Calculations were done at baseline (blue) conditions as well as during hyperoxia (red). (reproduced with permission from Werkmeister et al., 2012b).

Fig. 12

Sample phase extraction of a baseline measurement. Over the measurement period of 12 s the eye moved relative to the incoming laser beams resulting in pronounced changes in Φ₁ and Φ₂, but almost unchanged ΔΦ. (reproduced with permission from Werkmeister et al., 2012b).

Fig. 13

Relative error induced by an eye movement of the angle δ between −3 and +3° on Φm/Φ depending on the angle of incidence.

Fig. 14

Schematic diagram of a three-beam measurement geometry to extract the full Doppler angle: top- and side-view on the left, three-dimensional (3-D) model on the right. (reproduced with permission from Trasischker et al., 2013).

Fig. 15

Time course of blood flow velocity evolution in all vessels scanned by a rotating Dove prism. Red and blue represent the axial velocity measured for the two channels, left scale. It permits to calculate the absolute velocity (black line) and its mean value (dotted line), right scale. (e) is a ∼15° fundus view centered at the ONH obtained by calculating the en-face mean projection of an OCT 3D data set. (reproduced with permission from Blatter et al., 2013a).

Fig. 16

Scanning pattern around the optic nerve head. The scanning positions denoted by lines in the fundus image (left panel) and the corresponding OCT phase images for each scanning position (right panel) are presented.

Fig. 17

(a) Recording scheme for DOCT angiography; (b) Calculated 3D DOCT angiogram; (c) taking the en-face projection by e.g. plotting the maximum intensity results in comprehensive en-face angiography maps; (d) zoomed region of dotted box in (b) showing the signal decorrelation tails below vessels.

Fig. 18

Correction for phase-wrapping artifacts. (lhs) Phase tomogram with wrapping artifacts in positive phase shift direction in the vessel center. (rhs) Phase tomogram after unwrapping (reproduced with permission from Werkmeister et al., 2012a).

Fig. 19

50 × 50° wide field OCT angiography: (a) OCT intensity fundus projection; (b) tomogram cross section at indicated (white line) vertical position in (a). the colors denote the depth sections for which OCT angiogram projections are taken, and combined for the RGB representation in (c). (c) OCT angiogram with color-coded depth; boxes I and II indicate large and small choroidal vasculature, respectively, (green, blue) visible together with inner retinal vasculature (red). (reproduced with permission from Blatter et al., 2012).

Fig. 20

(a) OCT fundus projection of 12 × 12° Field of view. (b) OCT angiogram of the same patch. The projection is taken from the retinal nerve fiber layer down to the outer nuclear layer. (c) Projection from just below retinal pigment epithelium down to the choroid. The dense choriocapillaris network is well appreciated (reproduced from Blatter et al., 2012 with permission).

Fig. 21

(a) 3D rendered OCT angiogram of optic nerve head region, with an OCT fundus projection image below; (b) Zoomed OCT angiogram of lamina cribrosa architecture; Field of View 8 × 8°.

Fig. 22

Right macular images of a 45-year-old man with choroidal neovascularization secondary to myopia. (a) Color fundus image. (b) Fluorescein angiography (late-phase). (c) indocyanine green angiography (mid-phase). (d) En face projections of structural OCT. (e) En face projections of high-sensitive OCT angiography (reproduced from Hong et al., 2013 with permission).

Fig. 23

Left macular images of a 62-year-old man with age-related macular degeneration. (a) Color fundus image. (b) Fluorescein angiography (late-phase). (c) ICG angiography (mid-phase). (d) En face projection of structural OCT. (e) En face projections of high-sensitive OCT angiography (reproduced from Hong et al., 2013 with permission).