Application of OCTA in diabetic retinal microangiopathy

CHEN Ye‑yun, SONG Xu‑hua, CHEN Xiao‑yan, LI Lei

1.Department of Ophthalmology,the First Affiliated Hospital of Hainan Medical University,Haikou 570102,China

2.Hainan New Hope & Aier Eye Hospital,Haikou 570102,China

3.Affiliated Hainan General Hospital,Haikou 570102,China

Keywords:

ABSTRACT Diabetic retinopathy (DR) is the most common microvascular complication of diabetes mellitus, It can cause blindness without early intervention.Optical coherence tomography angiography (OCTA) has been featured by non‑invasiveness, fastness, safeness and high resolution, which can observe and analyze quantitatively the changes of retinal and choroidal vessels at different levels.In this paper, the basic principle of OCTA and its application in microaneurysms, intraretinal microvascular abnormalities, neovascularization, diabetic choroidopathy and diabetic neuronal injury of DR have been reviewed, to understand the different morphological changes of DR, and to realize the early screening, monitoring and evaluation of the treatment effect of DR.

1.Introduction

The increasing prevalence of diabetes mellitus (DM) is already a serious global health problem.The global prevalence of DM was about 9.3% (about 463 million people) in 2019 and it is expected to rise to 10.9% (about 700 million people) in 2045.About 116 million people in China have diabetes, which will rise to 147 million by 2045[1].And Diabetic Retinopathy (DR) is the most common complication of diabetes in the ocular microvasculature.Combining the epidemiological data of DR in China in the last three decades,the overall prevalence of DR in China was 1.14% and 18.45% in the population without diabetes and diabetes, respectively[2].DR is a major cause of blindness in working age population, with diabetic macular edema and neovascularization posing the greatest risk to vision.Effective prevention and management of diabetes is of great significance to reduce the rate of DR blindness and improve the quality of life of individuals and society.Optical Coherence Tomography Angiography (OCTA) is a non‑invasive, emerging vascular imaging technique that has developed rapidly in recent years, which is another major breakthrough after Optical Coherence Tomography (OCT).OCTA can detect and quantify retinal and choroidal blood flow at different levels in a very short period of time without contrast injection, and has a more intuitive and precise imaging effect on the extent and depth of ocular lesions.OCTA breaks through the limitations of Fundus Fluorescence Angiography(FFA) such as allergy, infection and long examination time caused by fluorescein injection.It makes up for the defect that FFA can only show the superficial blood vessels of the retina and only qualitatively assess the severity of DR.It has made three major leaps from FFA to OCTA in terms of qualitative to quantitative,invasive to non‑invasive, and two‑dimensional to three‑dimensional development.In this paper, we will review the basic principles of OCTA and its application in DR to understand the different morphological changes of DR and to achieve early screening,monitoring and assessment of the treatment effect of DR.

2.Overview and advantages of OCTA

OCTA is an OCT‑derived vascular imaging technique.Its principle is to obtain high‑resolution 3D vascular images by performing continuous B‑scans at the same site and detecting the contrast of motion between the flowing red blood cells and the surrounding “static” tissue[3].Moreover, Jia et al.introduced the Spectral Spectral Amplitude Decoherence Angiography (SSADA)technique to eliminate the effect of axial motion noise within the tissue, improve the signal‑to‑noise ratio, and clearly display retinal choroidal vascular lesions[4].OCTA blood flow images can be generated by both “maximum intensity projection” and “average intensity projection” methods[5].Studies have shown that maximum intensity projection is more sensitive to smaller vessels and weak blood flow, and that deep choroidal vessels are better scanned using the maximum intensity projection algorithm with high repeatability[6].

OCTA has the advantages of quick, safe, and non‑invasive with high resolution, which can observe the blood flow of retina and choroid layer by layer and clearly show the tiny non‑perfused area or neovascularization masked by the fluorescence of FFA leakage to achieve multi‑dimensional and accurate observation.Early changes in DR mostly occur before the peripheral parts of the retina.The advent of Wide Field OCTA (WF‑OCTA) can effectively visualize peripheral retinal lesions.It has been shown that WF‑OCTA scan of 6×6 mm can obtain about 11.3×11.3 mm square retinal vascular images, and scanning 12×12 mm can obtain about 16.7×16.7 mm square images[7,8].This is important for observing early peripheral lesions, non‑perfused areas of diabetic retinal disease.In addition,OCTA quantifies DR retinal vascular morphology and density, etc.by built‑in quantitative analysis indexes, and it creates possibilities for describing and quantifying DR vascular changes, studying pathogenesis, developing new treatments, and evaluating treatment effects.

3.Application of OCTA in DR

DR according to different fundus features can be divided into non‑obvious retinopathy (NDR), non‑proliferative DR (NPDR) and proliferative DR (PDR).NPDR according to severity can be divided into mild, moderate and severe.DR is one of the most common and in‑depth diseases in OCTA application research.OCTA can detect the fundus features of DR, including microaneurysm, capillary nonperfusion area, foveal avascular zone, diabetic macular edema,intraretinal microvascular anomalies and neovascularization.

3.1 OCTA Quantitative Metrics for morphological characteristics of DR

3.1.1 OCTA Quantitative Metrics for Foveal Avascular Zone

The central macular recess of the human eye has a region lacking retinal capillary perfusion, called the Foveal Avascular Zone (FAZ).Studies have shown that changes in the FAZ are closely related to visual function[9].Choi et al[10] found that the FAZ in diabetic patients was observed using OCTA and the results showed that the mean values of the longest diameter of the FAZ in the normal control, NDR, NPDR and PDR groups were 575 ± 146 µm, 696 ±153 µm, 813 ± 208 µm and 1150 ± 165 µm, respectively.Yashi et al[11] studied the changes in FAZ area in preclinical DR and found that the FAZ area in the superficial capillary layer (SCP) and deep capillary layer (DCP) was increased in patients with preclinical DR compared to normal controls.Krawitz et al[12] used OCTA to measure the roundness index and axial ratio of FAZ in diabetic patients and found that the roundness index and axial ratio changed in different periods of diabetes and were significantly associated with the progression of DR.The above studies suggest that both the size and shape of the FAZ start to change in the early stages of diabetes and may be a sensitive indicator for screening and monitoring early changes in DR.It has also been suggested that parameters related to FAZ size (e.g., diameter, area, etc.) have greater variability than those related to FAZ shape (e.g., roundness index, axial ratio, etc.),and there is a certain range of overlap between healthy individuals and diabetic patients[13], so parameters related to FAZ shape may be better parameters for monitoring FAZ in DR.

3.1.2 OCTA Quantitative Metrics for retinal microcirculation

Some studies have found that the vascular density (VD) of superficial and deep capillary layer and choroid of para‑macular fovea has been decreased in the early clinical stage of DR[14,15].The results showed that there were signs of retinal and choroid vascular damage in the early clinical stage of DR.Dupas et al.[16] selected 22 patients with type 1 diabetes.They found that the reduction in the vascular density of superficial, intermediate, and deep capillaries,especially in deep capillaries, was accompanied by a significant decrease in visual acuity.Tang et al.[17] also observed 334 patients with diabetes.They found that the decrease of blood vessel density in the deep capillary layer was significantly correlated with the progression of DR and the fell of visual acuity.It is suggested that the pathological changes of DR begin with the deep vascular plexus and gradually develop to the superficial vascular plexus.The visual acuity of diabetic patients mainly depends on the change of vascular density in the deep vascular layer.Only the decrease of vascular density in the deep vascular layer is enough to cause visual dysfunction.Therefore, longitudinal observation of the blood flow indexes of the deep vascular layer is of great significance for monitoring, tracking, and predicting the progress of DR.

3.1.3 OCTA Quantitative Metrics for microaneurysms

Microaneurysms (MA) are the earliest clinical lesion characteristic of DR, which are an important part of DR severity grading.In addition, microaneurysms are considered to be an indirect sign of retinal ischemia.Microaneurysms are mostly deep red dots with regular, small and rounded margins in fundus color photographs and mainly appear as homogeneous dotted hyperfluorescence in FFA.Histopathological studies confirm that microaneurysms mainly originate from the deep capillary plexus in the inner nuclear layer and its inner and outer edges, showing focal elevations and saccular and spindle‑shaped dilatations[18,19].Using OCTA, Borrelli et al.[20]found that microaneurysms could be present in both superficial and deep retinal vascular plexuses, but were mostly located in the deep vascular plexus and showed saccular, spindle‑shaped focal dilatation.This is similar to the findings of histopathology.A study on the issue of OCTA’s ability to detect microaneurysms showed that about 97.5% of FFA images detected microaneurysms, while the detection rates of 3×3mm and 4.5×4.5 mm swept-frequency OCTA images were 56% and 65.8%, respectively[21].Parravano et al.[22]classified microaneurysms into three classes of low, medium, and high reflectivity based on reflectivity on frequency‑domain OCT,and further found that the detection rate of microaneurysms with low reflectivity on OCTA (66.7%) was significantly lower than that of aneurysms with medium and high reflectivity (88.9%).This phenomenon did not correlate with microaneurysm size, which may be related to turbulent or low flow velocity (less than 0.3 mm/s) of microaneurysm blood flow or aneurysms containing only plasma and no flowing red blood cells[23-25].In addition, microaneurysms detected by OCTA may also be capillary ends or capillaries distributed perpendicular to the retina.It indicates that although OCTA can adequately describe microaneurysm morphology and precisely localize them, there is still some deficiency in its detection ability.

3.1.4 OCTA Quantitative Metrics for Intraretinal Microvascular Anomalies

Intraretinal Microvascular Anomalies (IRMA) is one of the features of severe NPDR and is manifested in fundus color photography and FFA as tortuous, dilated vessels, shunted abnormal vessels and intra‑retinal neovascularization in the retina[26-28].Microvascular abnormalities and neovascularization within the retina can be clinically distinguished by FFA, but capillaries with abnormally dilated walls and abnormal vessels other than neovascularization may also have leakage, while neovascularization with minimal leakage may also be missed.Therefore, relying on FFA leakage alone is still not sufficient to identify the two.Compared with fundus color photography and FFA, OCTA has a higher detection rate and is more accurate for intraretinal microvascular abnormalities.This may be related to the fast flow rate of the abnormal vessels.Intraretinal microvascular anomalies appear on OCTA as abnormal vessels at the venous end of the retina originating from the edge of the nonperfused zone in the inner plexiform layer, ganglion cell layer, and nerve fiber layer, characterized by hyperreflective signals and localized inner border membrane (ILM) elevation within the retina[29].It can be dilated, ring‑shaped, pigtailed, reticulated and sea fan‑shaped, and intraretinal microvascular anomalies with a sea fan shape are more likely to break through the ILM and develop into neovascularization[30,31].Intraretinal microvascular abnormalities can further develop into neovascularization, and the presence of intraretinal microvascular abnormalities should be a warning for the risk of progression to PDR when found.

3.1.5 OCTA Quantitative Metrics for neovascularization

Neovascularization (NV) represents the main feature of PDR.Prolonged retinal hypoxia can lead to neovascularization, which can break through the inner boundary membrane and extend growth into the vitreous cavity.The specificity and sensitivity of wide‑field OCTA for detecting neovascularization are 97% and 100%,respectively, which is superior to the detection rate of Ultra Wide Field Fundus Fluorescence Angiography (UWFFA)[32].OCTA detects neovascularization by observing the blood flow signal in the inner boundary membrane to clarify its origin, location and vascular morphology.Carlo et al.detected neovascularization by using OCTA observation adjacent to the nonperfused area and at microvascular abnormalities within the retina[33].Neovascularization of the optic disc (NVD) originated from retinal arteries, veins, posterior ciliary arteries and choroidal vessels at the edge of the optic papilla[34].Khalid et al.[35] combined OCTA en face images of the vitreoretinal junction with tomographic images and found NVD to be highly reflective tissue protruding from the retinal surface at the vitreoretinal junction near the optic disc.Pan et al.[30] observed the eyes with PDR using OCTA and classified the neovascularization into three types according to its origin and branching morphology.Type 1 (42.67%)originates from the venous system at the edge of the non‑perfused area of the ganglion cell layer and nerve fiber layer, shaped like a dendrite; Type 2 (40%) originates from the capillary plexus within the nonperfused zone of the inner nerve layer and breaks through the inner boundary membrane in multiple directions, shaped like an octopus; Type 3 (17.33%) originates from abnormal microvessels in the nonperfused zone of the inner nuclear layer and ganglion cell layer, shaped like a sea fan, which can break through the inner boundary membrane and grow towards the vitreous cavity.Vaz‑Pereira et al.used OCTA to study the cross‑sectional morphology of neovascularization and classified it into flat, forward elevated,or table‑top morphology, and most of the neovascularization was predominantly forward elevated.The flat and tabletop shapes did not break through the posterior vitreous membrane, while the forward bulge shape broke through the posterior vitreous membrane and grew forward[36,37].OCTA can observe the origin and morphology of PDR neovascularization, which can help provide new directions in the classification of retinal neovascularization and better understand the pathophysiological mechanisms of DR.

3.2 OCTA for evaluating effect of DR

The morphological characteristics of DR on OCTA have been described previously.OCTA can also be used to observe and evaluate the response, efficacy and prognosis of DR after intravitreal injection of anti‑vascular endothelial growth factor (VEGF) drugs, steroid hormones and retinal laser photocoagulation.

3.2.1 OCTA for evaluating effect of DME

The size and morphology of microaneurysms are not invariable.Regression is considered to be a sign of good prognosis for DR[38].The relationship between microaneurysms and diabetic macular edema (DME) has been well documented in a number of studies.Studies have confirmed that in patients with diabetic macular edema, the majority of microaneurysms are present near the saccular cavity, with only 10‑20% protruding into the cavity[39].Pongsachareonnont et al.[40] observed changes in macular edema before and after anti‑VEGF treatment in diabetic patients and showed that before treatment, the total number of microaneurysms in the superficial and deep capillary layers was 5.24 ± 2.67 and 18.72± 4.04, respectively, and the central macular thickness (CMT) was 397.64 ± 91.17 µm; At 1 month after treatment, the total number of microaneurysms in the superficial and deep capillary layers was 2.26±1.90 and 12.67±3.84, respectively, and the central macular thickness was 364.59±84.3 µm.These data showed a significant reduction in the number of microaneurysms and a reduction in macular edema after anti‑VEGF treatment.Nevertheless, after observing the characteristics of microaneurysms and performing a one‑year follow‑up, Parravano et al.[41] found that OCTA observed a greater number of microaneurysms in the deep vascular layer of the macula than in the superficial vascular layer; Microaneurysms with high reflective signal have a fast flow rate and may cause a vascular inflammatory response that is involved in the process of blood‑retinal barrier damage and increased vascular permeability and accumulation of extracellular fluid in the retina, which leads to or exacerbates increased retinal thickness and edema, and it is believed that their increased number may cause DME and promote the progression of DME; Moreover, microaneurysms with high reflective signals respond well to anti‑VEGF and steroid hormone therapy and may reduce the degree of edema in DME.In contrast,microaneurysms with low reflectivity are less effective against VEGF and steroid hormone therapy because they rarely cause the above‑mentioned inflammatory response and the process of retinal damage and edema.Thus the characteristics of microaneurysms can be used as biomarkers to predict the response to anti‑VEGF therapy,the severity of macular edema, and the progression of DR.

3.2.2 OCTA for evaluating effect of IRMA

Shimouchi et al.[42] showed morphological changes of microvascular abnormalities within the retina before and after total retinal laser photocoagulation (PRP) comparing severe NPDR and PDR.Five types were described as (1) Immutable, (2) Cluster regression, (3) Reperfusion, (4) Deterioration, and (5) Mixed, and the immutable and reperfusion types indicated some remodeling ability of abnormal microvessels in the retina.Similarly, Sorour et al.[31]observed the regression of intraretinal microvascular abnormalities in eyes with PDR that had received anti‑VEGF treatment.Observations using OCTA revealed that some of the abnormal microvascular branches normalized and merged into the surrounding capillary bed, and the nonperfused area was reduced, but there was also an increase in the number of abnormal microvascular branches or occlusion and an increase in the extent of the nonperfused area.They also summarized the regression of these abnormal microvessels into four conditions: improving (31%), remaining unchanged(44%), progressing (13%), and completely occluded (11%).All of these studies suggest that whether retinal laser photocoagulation or anti‑vascular endothelial growth factor treatment is taken,abnormal microvessels in the retina have the potential to convert to normal vessels after treatment, which has positive implications for preventing their further development.

3.2.3 OCTA for evaluating effect of PDR

The mechanism of retinal ischemia induced by vascular endothelial growth factor (VEGF) remains unclear.It has been reported that the upregulation of vascular endothelial growth factor leads to an increase in the expression of intercellular adhesion factor‑1 in endothelial cells, followed by increased adhesion of activated monocytes and granulocytes to endothelial cells, resulting in capillary obstruction.The more vascular endothelial growth factor is released after retinal ischemia, forming a vicious circle of aggravation of retinal ischemia and enlargement of the non‑perfusion area[43-45].Alagorie et al.[46] observed the changes in macular vascular density in patients with PDR after the vitreous cavity injection of anti‑VEGF drugs.It was found that the vascular density of the superficial and deep retinal vascular network and choroid plexus was not significantly lower than that before treatment.Therefore, it is speculated that anti‑VEGF drugs will not lead to further aggravation of retinal ischemia and progressive enlargement of the perfusion area, which plays a positive role in treating PDR.Figueira et al.[47] used OCTA to compare the efficacy of ranibizumab plus PRP and PRP alone in treating PDR 12 for months and found that about 43.9% of the eyes in the ranibizumab group had complete neovascularization, compared with 25.0% in the PRP group alone.HE et al.[48] also compared the efficacy of PRP alone and PRP+ conbercept in treating PDR 6 months later.The results showed that compared with those treatments before, the size and area of neovascularization were significantly reduced, the best‑corrected visual acuity was improved to some extent in both groups, and the change in the combined treatment group was more significant than that in the PRP group alone.Both groups of studies have shown that retinal laser photocoagulation combined with anti‑VEGF drugs or retinal laser photocoagulation alone can effectively promote the regression of neovascularization and control PDR progression.The effect of combined therapy is better than that of retinal laser photocoagulation alone, which can reduce the times of injection of anti‑VEGF drugs and the times and energy of retinal laser photocoagulation.Therefore, the combination therapy can be used for patients with proliferative diabetic retinopathy.

4.OCTA Quantitative Metrics for diabetic choroidopathy

Choroid is a vascular‑abundant tissue in the eye with abundant blood flow.It is divided into choroid capillary layer (CC), Sattler vascular layer and Haller vascular layer.The choroid cycle is an important source of oxygen and nutrients for the choroid and the outer layer of the retina (including the pigment epithelium and photoreceptors).Diabetic choroidopathy (DC) refers to the degeneration of choroid capillaries, microaneurysms, vascular tortuosity, verrucous deposition on Bruch membrane and choroidous neovascularization caused by diabetes[49].Although indocyanine green angiography (ICGA) is the “gold standard” for the observation of choroid lesions,but due to its own limitations, it is still unable to accurately observe the subtle changes in choroid blood flow, the emergence of OCTA makes it possible.

OCTA grasps the blood flow of DC by observing the capillary blood flow density (CFD), vascular density (VD), and the area of flow deficit (FD).Yang et al.[50] used OCTA to observe the choroidal blood vessels in DR patients’ 3x3mm macular area.The results showed that the choroidal capillary blood flow density values of the control group, no DR group, mild NPDR group, moderate NPDR group, severe NPDR group, and PDR group were 65.66(0.50), 63.50 (0.61), 62.07 (0.67) and 58.13 (1.04), respectively.The mean values of choroidal capillary blood flow density in the macular edema‑free and macular edema groups were 62.39(0.58)and 53.98(0.60).It is suggested that there are signs of choroidal blood flow decline in the early stage of DR, and this sign becomes more and more serious with the progress of DR, especially in DME patients.This conclusion is still supported even if the observation range is extended to 6x6mm.Dai et al.[51], using OCTA, found that compared with the normal control group, patients without DR had reduced choroidal blood perfusion in the macular area.It is suggested that choroid perfusion may be an early detection index for patients without DR.Ra et al.[52] further studied the vascular density of retina and choroid by OCTA and found that with the development of DR, the vascular density of the superficial and deep retinal vascular network and choroidal capillaries decreased, while the vascular density of large choroidal blood vessels increased.The results showed that the severity of DR had little influence on the large choroidal blood vessels.It was speculated that the increased vascular density of large choroid blood vessels might be the compensation mechanism for the decreased vascular density of retina and choroid capillaries in diabetic patients.

The outer structure of the retina is mainly supplied by choroidal circulation, especially the macular ellipsoid zone.Damage to choroidal microcirculation may affect visual function.Ro‑Mase et al.[53] used OCTA to observe the relationship between the changes of choroidal blood flow and visual function in the macular area of DR.The results showed that there was a significant correlation between the increase of macular insufficient blood flow area and the decrease of retinal sensitivity in DR group compared with the healthy control group and non‑DR group.It is suggested that choroidal ischemia in DR is more likely to cause macular visual dysfunction.In summary, OCTA has been proved to be a new and effective detection method for studying choroidal blood flow and function and is of great significance in exploring the choroid’s pathophysiological mechanism in DR.

5.OCTA Quantitative Metrics for diabetic neurodegeneration

Diabetic retinopathy is not simply microangiopathy but is related to oxidative stress, elevated glutamate levels, activation of Müller cells and overexpression of glial cells in the renin‑angiotensin system under chronic hyperglycemia[54].

Microvascular injury of DR has been extensively reported in many studies.It has been confirmed that vascular density in all retinal layers decreases with increasing severity of DR.However,the role of nerve injury, degeneration, and neuroinflammation in the development of diabetes is still being explored.Diabetic neurodegeneration is a progressive and degenerative process of the retinal nerve cell layer, and ganglion cell apoptosis has appeared in the early stage.Histologically, neuronal apoptosis and glial cell activation were observed in the ganglion cell layer[55].OCTA can observe changes in microcirculation and between neurons in areas such as the macula.Kim et al.used OCTA to detect 155 DR patients.They found that the vascular density and perfusion index in the macular area of NDR and NPDR patients decreased compared with healthy people, which was related to the thinning of the ganglion cell layer and plexiform layer (mGC/IPL)[56].It can be seen that the damage of macular microvessels and neurons has occurred in the early stage of diabetes.However, it has been pointed out that the importance of microvascular changes in radial peripapillary capillary plexus (RPCP) in DR has not been extensively studied so far due to the difficulty of imaging[57].

The radial capillary plexus around the optic disc is composed of the superficial capillaries located in the inner layer of the optic nerve, which is mainly supplied by the central retinal artery.After passing through the center of the optic papilla, the central retinal artery is continuously divided into small branches to form a five‑layer vascular structure in the retina.The vascular structure around the optic papilla is the thickest and most compact[58].It is speculated that when the vascular density decreases in the macular area at the end of the blood supply, the vascular density of the optic papilla is also likely to decrease.Cao et al[59].found that the vascular density in and around the optic papilla of diabetic patients decreased significantly compared with normal subjects.At the same time, the thickness of the retinal nerve fiber layer (RNFL) became thinner in nasal and superior quadrants.Li HD et al.[60] also found that the blood flow density of RPCP decreased significantly in the whole week and each quadrant in the diabetic group, while the thickness of RNFL decreased more significantly in the areas above and below.It is suggested that there are signs of optic papilla microcirculation and neuropathy in the preclinical stage of DR.The changes of optic papilla microcirculation maybe earlier than that of the nerve fiber layer, and both of them may have regional changes.It is speculated that there is no direct correlation between the changes of optic papilla blood flow and the loss of optic nerve fibers, and nerve injury can not be regarded as secondary to retinal microvascular injury.Although neurons and vascular structures are functionally related,there may be different pathological mechanisms.These mechanisms may lead to retinal vascular injury and neuronal loss in different ways and times.In conclusion, the exact relationship between microvascular injury and optic papillary neurons in DR has not been determined, and its early pathogenesis has not been explained.Therefore, therapeutic strategies targeting the signaling pathways leading to microvascular dysfunction may be of great significance for preventing the development of DR and neurodegeneration.

6.Summary and Prospect

OCTA as a non‑invasive angiography technique.Due to its qualitative and quantitative characteristics of retinal blood vessels,compared with FFA and ICGA, it can more accurately visualize and quantify retinal and choroidal blood vessels, which is conducive to the transverse and longitudinal observation of retinal diseases.At present, it is widely used in the study of vascular morphology and blood flow changes in diseases such as diabetic retinopathy, retinal vein occlusion, age‑related macular degeneration, cystoid macular edema, choroidal neovascularization, etc.The detection rate of retinal vascular abnormalities and neovascularization by OCTA is better than that by traditional methods, which is helpful to improve the detection rate, diagnosis, and treatment rate of PDR.The size and shape of FAZ, circulatory density, vascular density, and other quantitative indicators can also be examined to detect early clinical DR lesions.However, the common problems of OCTA inevitably exist.First of all, the limitation of the observation field, severe opacity of turbid media, limitation of various artifacts (projection artifacts, masking artifacts, motion artifacts), segmentation errors,and other issues can affect the clarity of the image.Secondly,OCTA imaging is affected by the decline of blood flow rate and resolution as the imaging range expands, leading to a low detection rate of microaneurysms and overestimation of the non‑perfusion area[61].With the development of the Times, various technologies such as montage technology, artifact removal technology, eye movement tracking technology emerge in an endless stream, which not only reduces the requirements of patients with fixation but also improves the quality of OCTA images to better observe the changes of retinal, choroidal vessels.In conclusion, although OCTA still has shortcomings at the present stage, we expect it to continue developing, breaking through the shackles, and providing better detection methods for diagnosing and treating ophthalmic diseases and follow‑up.

Contribution description:

Chen Yeyun.Reviewing and summarizing the references.Writing full text.

Chen Xiaoyan: Technical guidancing.

Song Xuhua: providing scientific research project supporting,technical guidancing.

Li Lei: Technical guidancing.

None of the above authors has a conflict of interest.