OCTA detects change between repeated OCT measurements at the same retinal location and uses that change as a proxy for moving blood. It can separate vascular layers by depth, but the final angiogram depends on acquisition quality, segmentation and flow-detection thresholds.
- OCTA does not inject dye and does not directly photograph blood; it derives vascular contrast computationally from repeated scans.
- Depth-resolved slabs are a core strength, but automated slab boundaries can fail when disease distorts normal anatomy.
- Absent OCTA signal can mean nonperfusion—or slow flow, shadowing, poor signal or processing failure.
The basic idea: look for change
If the same retinal location is scanned repeatedly, stationary tissue should produce a relatively stable signal while moving blood cells create measurable change. OCTA algorithms exploit that difference. The result is a three-dimensional flow-related dataset that can be projected into en-face maps. Early methods such as split-spectrum amplitude-decorrelation angiography demonstrated that decorrelation could improve visualization of retinal and optic-nerve-head microvasculature. Modern commercial systems differ in implementation, but the central concept remains motion contrast.
Why depth resolution matters
Fluorescein angiography produces a two-dimensional view of fluorescence through time. OCTA begins with depth-resolved OCT data, allowing software to isolate vascular slabs. That means superficial and deeper retinal plexuses can be examined separately, and choriocapillaris or neovascular networks can be displayed in distinct planes. This is one of OCTA's most important conceptual advantages: vascular information is tied to anatomical depth rather than collapsed into a single projection.
How an en-face angiogram is built
A scanner acquires repeated B-scans at closely spaced positions. Software compares the repeated measurements, computes a motion-related metric and reconstructs a volume. Segmentation boundaries then define the slab being displayed. The bright branching lines in an OCTA image therefore represent detected flow signal within a selected depth range. They are not a literal photograph of vessels. Change the slab boundary, threshold or algorithm and the visible network can change.
What OCTA can reveal
OCTA is useful for visualizing capillary networks, areas of reduced detectable flow, the foveal avascular zone and neovascular complexes. Because there is no dye leakage to blur vessel borders, vascular morphology can be strikingly clear. In some neovascular disorders, OCTA can reveal a vascular network that is difficult to distinguish on fluorescein angiography. It can also be repeated without intravenous contrast, which makes serial structural-plus-vascular assessment practical.
The slow-flow problem
A dark region on OCTA is not automatically proof that a vessel is absent. Flow below the algorithm's detection threshold may produce little or no signal. Shadowing from hemorrhage, exudate or other structures can also reduce signal. Conversely, motion from the eye or scanner can create false vascular patterns. Interpreting OCTA therefore requires a distinction between 'no detected flow signal' and 'no blood flow.' Those statements are not equivalent.
Segmentation is both power and vulnerability
Depth-resolved analysis depends on accurate segmentation. Healthy anatomy is relatively easy for automated software to model. Pathology can be much harder: fluid, atrophy, pigment-epithelium detachments and distorted retinal layers can pull automated boundaries away from the structures they were intended to follow. Artifact studies have found substantially more segmentation errors in diseased than healthy eyes. Reviewing the B-scan with flow overlay is therefore part of responsible interpretation.
Why OCTA does not show leakage
OCTA detects motion-related contrast, not fluorescent dye. A leaky vessel may be visible as a vascular structure, but the surrounding leakage that fluorescein angiography demonstrates is not directly represented. That is not a minor technical omission; leakage can be clinically meaningful. OCTA is therefore best understood as adding a different vascular dimension, not as a universal replacement for dye angiography.
What better OCTA would look like
The frontier is moving toward wider fields, faster acquisition, improved motion correction, more reliable segmentation and quantitative biomarkers that generalize across devices. The challenge is not merely generating prettier angiograms. It is making measurements reproducible enough that vessel density, nonperfusion and neovascular morphology can be compared across time, scanners and populations without confusing algorithmic variation for biological change.
Quantifying vessels is harder than drawing them
Once a vascular map exists, software can convert it into metrics such as vessel density, perfusion density, skeletonized vessel length or nonperfusion area. The apparent objectivity can be deceptive. Thresholding decisions determine which weak signals count as vessels, and different scan sizes change sampling density. A metric that is reproducible on one device may not match the same named metric on another. Quantitative OCTA is therefore most persuasive when acquisition, processing and test-retest behavior are explicitly documented.
The scan-size tradeoff
Small OCTA fields can sample the macula densely and reveal fine capillary detail, while larger fields cover more retina at the cost of lower sampling density or longer acquisition. Modern high-speed systems partly relax this tradeoff, but do not eliminate it. When comparing studies, field size is not a footnote: a 3×3-mm scan and a widefield montage may answer different questions and produce different vessel-density values simply because they sample different anatomy at different resolution.
OCTA implementation is device- and algorithm-dependent. Quantitative values should not be assumed interchangeable across platforms without validation.
Sources & original records
We prioritize primary records, clinical-trial registries, peer-reviewed literature and authoritative institutions. Manufacturer material is labeled when used to describe a product or company position.
- Split-spectrum amplitude-decorrelation angiography with optical coherence tomographyPubMed · Primary methods study · PMID 22418228
- Prevalences of segmentation errors and motion artifacts in OCT-angiography differ among retinal diseasesPubMed · Artifact study · PMID 29982897 · DOI 10.1007/s00417-018-4053-2
- Clinical Features Related to OCT Angiography Artifacts in Patients with Diabetic Macular EdemaPubMed · 2024 · Clinical artifact study · PMID 38447922