Ophthopedia flags OCTA standardization gaps
- Ophthopedia highlighted a 2026 AMD imaging review arguing OCT angiography still lacks standardized ways to measure and describe macular neovascularization across studies. - The gap is concrete: papers use different scan sizes, slab segmentations, lesion-area definitions, and “activity” markers, so results often cannot be compared cleanly. - That matters because OCTA is moving toward treatment monitoring, but consensus rules are still catching up to clinical ambition.
OCT angiography is one of the most promising tools in retina care because it can show abnormal blood vessels in wet AMD without injecting dye. That is a big deal. It is faster, safer, and easier to repeat over time. But the catch is that seeing more detail is not the same as measuring it the same way every time. That is the gap this latest review puts front and center — the field has a powerful imaging tool, but not yet a shared rulebook for how to quantify macular neovascularization, or MNV. ### What is OCTA actually measuring? OCTA tracks motion from blood cells and turns that into a map of retinal and choroidal blood flow. In neovascular AMD, the thing clinicians care about is the abnormal vessel network under or within the retina — the MNV lesion. OCTA can show that network in segmented layers, which is why people hope it can do more than just detect disease and start helping with classification, monitoring, and maybe treatment timing. (nature.com) ### So where does the standardization problem start? It starts almost immediately — at image acquisition. Different studies use different devices, scan sizes, and processing pipelines. A 3×3 mm scan does not capture the same lesion detail as a 6×6 mm scan. Spectral-domain and swept-source systems also behave differently. Then come artifacts, projection errors, and signal-quality issues, all of which can change what the lesion looks like before anyone even starts measuring it. (nature.com) ### Why can’t researchers just compare lesion size? Because “lesion size” is not one thing. Some papers measure the whole MNV complex. Others isolate perfused area, vessel density, total vessel length, junction density, lacunarity, or flow deficit patterns around the lesion. Even when two groups say they measured area, they may have used different segmentation slabs or different thresholding rules to decide what counts as vessel and what counts as background. Basically, the same eye can produce different numbers depending on the recipe. (mdpi.com) ### What about “activity” markers? That is even messier. Researchers have proposed features like a dark halo, branching capillaries, loops, peripheral arcades, vessel shape, and anastomoses as signs of active disease. But the field still disagrees on which of these markers really tracks exudation, treatment response, or future reactivation. Some studies suggest OCTA features are useful. Others show those features may reflect prior anti-VEGF treatment as much as current activity. (nature.com) ### Haven’t experts tried to fix the terminology? Yes — and that is important context. An international expert panel already warned that OCTA terminology around neovascular AMD had become inconsistent, and consensus work has continued into 2026 with practical AMD imaging guidelines. So this is not a new complaint. What is new is how urgent it feels now that OCTA is being pushed toward biomarker development, AI models, and treatment monitoring. (nature.com) ### Why does this matter for AI? AI loves consistent inputs. If training data come from studies that define lesions, slabs, and activity in different ways, models can look impressive but learn a moving target. Recent reviews are optimistic about AI in OCTA, but they keep landing on the same bottleneck — data quality and harmonized labels. Without that, automation can scale inconsistency instead of solving it. ### Does this mean OCTA is not clinically useful? (aaojournal.org) No. It means OCTA is useful, but the quantitative layer still needs guardrails. Clinicians can absolutely use it alongside structural OCT and, when needed, dye angiography. The warning is narrower: do not overread a single OCTA metric as if the field has already agreed on its exact meaning or treatment implication. ### Bottom line? OCTA in wet AMD is ahead on capability and behind on standardization. (nature.com) The technology is ready for bigger clinical roles. The vocabulary, segmentation rules, and measurement conventions still are not. Until those pieces settle, the smartest use of OCTA is as a powerful adjunct — not a standalone scoreboard. (mdpi.com)
