Optical coherence tomography based angiography [Invited]

Abstract

Optical coherence tomography (OCT)-based angiography (OCTA) provides in vivo, three-dimensional vascular information by the use of flowing red blood cells as intrinsic contrast agents, enabling the visualization of functional vessel networks within microcirculatory tissue beds non-invasively, without a need of dye injection. Because of these attributes, OCTA has been rapidly translated to clinical ophthalmology within a short period of time in the development. Various OCTA algorithms have been developed to detect the functional micro-vasculatures in vivo by utilizing different components of OCT signals, including phase-signal-based OCTA, intensity-signal-based OCTA and complex-signal-based OCTA. All these algorithms have shown, in one way or another, their clinical values in revealing micro-vasculatures in biological tissues in vivo, identifying abnormal vascular networks or vessel impairment zones in retinal and skin pathologies, detecting vessel patterns and angiogenesis in eyes with age-related macular degeneration and in skin and brain with tumors, and monitoring responses to hypoxia in the brain tissue. The purpose of this paper is to provide a technical oriented overview of the OCTA developments and their potential pre-clinical and clinical applications, and to shed some lights on its future perspectives. Because of its clinical translation to ophthalmology, this review intentionally places a slightly more weight on ophthalmic OCT angiography.

Keywords: (110.4500) Optical coherence tomography; (170.2655) Functional monitoring and imaging; (170.3880) Medical and biological imaging.

Figures

Fig. 1

Number of optical coherence tomography angiography publication by year since 2004. Data source: PubMed (https://www.ncbi.nlm.nih.gov/pubmed), with “optical coherence tomography angiography” and “Doppler OCT” as the search key words. Data retrieved on October 30, 2016.

Fig. 2

A simplified schematic figure of the concept of optical coherence tomography based angiography. Signals are sampled from five points in the A-scan, where three pixels (1, 2, and 5) are located at the static tissue, and two pixels (3 and 4) are located within a functional blood vessel. Dynamic changes in the OCT signals for pixel 3 and 4 can be observed over time while signals from pixel 1, 2, and 5 remain steady.

Fig. 3

Example of OCTA angiograms of retina from a normal subject. (A) en face structure image, (B) cross-sectional OCT image sampled along the orange dotted line in (A); (C), (D), and (E) angiograms from superficial retinal layer, deep retinal layer, and avascular retinal layer, and (F) angiogram from the whole retinal layer with depth information encoded in false color.

Fig. 4

Flow quantification simulation results of OMAG signal intensity of (A) various B-scan time interval with multiple velocity scale and a magnify view of a red box in (A) indicating the flow velocity between 0 to 1.5 mm/s, and of (B) various particle concentration.

Fig. 5

Top row: OCTA angiograms using OMAG method from a 33 year old man diagnosed with proliferative diabetic retinopathy (PDR). (A) early phase fluorescein angiography image. (B) The defects observed on the OCTA image (projected within the superficial retinal layer, size: ~12 mm x 12 mm) that correspond to FA image, as indicated by the red arrows and green ovals. (C) Zoom-in to a 3 mm x 3 mm area centered at the foveolar from (B). Bottom row: OCTA images of 3 mm x 3 mm from a polypoidal choroidal vasculopathy (PCV) eye. (D) OCTA image from OMAG method projected within avascular retinal layer where projection artifacts are prominent that would affect the interpretation of the results, (E) OCT structural cross-sectional image sampled along the yellow dotted line in (D), and is superimposed with flow signal with color-encoded depth information. (F) Vasculature in the avascular retinal layer after projection artifact removal, clearly showing the choroidal neovascularization within polyps and branching vascular network next to it in the diseased eye.

Fig. 6

Images from the inflammation, proliferation, and maturation stages of wound healing over 10 days [125]. (A)-(C) OCT cross-sectional images sampling along the dashed lines in the en face images (D)-(F). (G)-(I) OMAG images projected within 1 mm depth. (J)-(L) Overlay of (D)-(F) with (G)-(I).

Fig. 7

(A) and (B) OCTA images of mouse cerebral cortex through skull using OMAG [138]. (C) OCTA image of mouse brain bearing a human glioblastoma tumor imaged with phase-based OCTA through a cranial window. Image was projected within the first 2 mm. Depth is encoded by color: yellow (superficial) to red (deep) [150]. (D) OCTA images using a high-pass filtered intensity-based OCTA through a cranial window [139]. (E) Volumetric OCT angiography imaging of the rodent cortex during ischemic stroke (1) at baseline, (2) progressive focal ischemia developed during middle cerebral artery occlusion (MCAO), and (3) 30 minutes after onset of reperfusion [151].