Author: Claude (AI), from PubMed abstracts · Reviewed by: Dr Mina Abdelmseih · Reviewed: 2026-11-15 · Review due: 2027-11-15

Educational content only. Not personalised medical advice. Clinical decisions remain the responsibility of the treating clinician.

OCT angiography

OCT angiography (OCTA) is a signal processing and scan acquisition approach that lets OCT devices show retinal and choroidal vessels down to capillary level without dye [S10]. Unlike dye-based angiography it avoids risks such as anaphylaxis [S11]. It detects choroidal neovascularisation (CNV) with high accuracy [S1] and is used in diabetic retinopathy, glaucoma, geographic atrophy and neuro-ophthalmology [S2] [S6] [S7] [S8], but artefacts are common and are its main clinical limitation [S3].

Key points - A meta-analysis of 16 studies (447 CNV eyes, 414 controls) found pooled sensitivity 0.87 and specificity 0.97 for detecting CNV [S1]. - Artefacts from eye motion, layer misidentification (segmentation), projection and low signal are the most common; recognising them is essential for correct diagnosis [S3]. - In diabetic retinopathy OCTA measures the foveal avascular zone, vessel density and neovascularisation; poor-quality scans are relatively frequent and re-imaging is often needed [S7]. - Conventional OCTA has a small field of view; widefield OCTA now reaches the mid-periphery and beyond [S10] [S11]. - In glaucoma, OCTA has good repeatability and can extend monitoring of advanced disease beyond the "floor effect" of structural OCT [S8].

Evidence last checked: 2026-10-17 (PubMed abstracts).

Red flags: findings that need urgent action

• A neovascular network on OCTA in a symptomatic eye. OCTA has high diagnostic value for CNV [S1]; confirm with structural OCT and refer for treatment [G1].
• Neovascularisation on widefield OCTA in diabetic retinopathy. OCTA can identify neovascularisation [S7]; refer to retina [G2].
• CNV in an eye with uveitis. Inflammatory CNV is the third leading cause of CNV and is hard to separate from other uveitic lesions; OCT and OCTA help early detection [S9].
• An "abnormal" OCTA on a poor-quality scan. Check for artefact before acting [S3].

Recognise it: what OCTA shows

• Layers (slabs). OCTA visualises capillaries in all retinal layers in three dimensions [S7]; a full-retinal slab limits segmentation errors and confounding from centre-involved diabetic macular oedema [S7].
• Diabetic macular ischaemia. Traditionally graded by an enlarged, irregular foveal avascular zone, but these changes are not surrogate markers of visual impairment; OCTA perfusion nomenclature still needs standardising [S4].
• Geographic atrophy. Choriocapillaris flow changes are associated with GA development, extend beyond atrophy borders, are prognostic for progression and may help separate GA from mimics [S2].
• Neuro-ophthalmology. Peripapillary and macular OCTA may aid early detection and differential diagnosis, and the detection of complications in optic disc drusen [S6].

Confirm it: is the image real?

1. Check scan quality. Proper scan quality evaluation is crucial for reliable metrics [S8].
2. Look for artefacts: motion, segmentation errors, projection and low signal [S3].
3. Device dependence. OCTA metrics vary with machine, post-processing algorithms, systemic disease and ocular factors [S8]; do not compare numbers across devices.
4. Correlate with structural OCT and other imaging; combining modalities supports personalised care in diabetic retinopathy [S5].

Manage it: using OCTA at each resource tier

Before acting on this section, check the national and local guidelines, product information and regulatory approvals that apply where you practise. See the medical disclaimer.

Tier A (primary care / no specialist)

No OCTA role. Refer symptomatic macular disease according to the table below.

Tier B (general eye clinic)

1. Where available, OCTA gives vascular imaging without dye risks such as anaphylaxis [S11].
2. Use OCTA alongside colour photography, OCT and, when needed, fluorescein angiography to evaluate diabetic retinopathy [S5].
3. Treat each OCTA finding as provisional until artefacts are excluded [S3].

Tier C (specialist centre)

1. CNV diagnosis and follow-up. Use OCTA to detect CNV [S1] and to separate inflammatory CNV from other uveitic lesions [S9].
2. Diabetic retinopathy. Use widefield OCTA, since peripheral vasculature matters for predicting progression [S7]. Allow time to re-image poor-quality scans [S7].
3. Glaucoma. OCTA differentiates glaucomatous from healthy eyes and correlates with structural and field changes; it complements but does not replace standard tests [S8].
4. Quantification and AI. Improved software and deep learning now allow precise quantification and automated diagnosis [S10]; clinical validation is still needed [S8].

Refer when

Pitfalls

• Mistaking projection or segmentation artefacts for disease [S3].
• Comparing vessel density values between devices [S8].
• Missing peripheral disease on small-field scans [S11].
• Equating an enlarged foveal avascular zone with visual loss [S4].
• Overcalling glaucoma from OCTA metrics alone [S8].

Evidence notes

• Only one source is a meta-analysis [S1]; the others are narrative reviews, several by technology developers.
• Evidence on inflammatory CNV rests on case reports and series, with no RCTs [S9].
• Not covered by these sources: accuracy of OCTA versus fluorescein angiography for retinal neovascularisation, numeric normal values, cost, and use in low-resource settings.

Related articles

• OCT interpretation: macula and retinal nerve fibre layer
• Fluorescein and indocyanine green angiography

References

See the numbered source list below. Each marker [S#] links to its PubMed and DOI record.