Diabetic Retinopathy Screening Research Articles

HbA1c changes in a deprived population who followed or not a diabetes self-management programme, organised in a multi-professional primary care practice: a historical cohort study on 207 patients between 2017 and 2019

BackgroundDiabetes self-management (DSM) helps people with diabetes to become actors in their disease. Deprived populations are particularly affected by diabetes and are less likely to have access to these programmes. DSM implementation in primary care, particularly in a multi-professional primary care practice (MPCP), is a valuable strategy to promote care access for these populations. In Rennes (Western France), a DSM programme was designed by a MPCP in a socio-economically deprived area. The study objective was to compare diabetes control in people who followed or not this DSM programme.MethodThe historical cohort of patients who participated in the DSM programme at the MPCP between 2017 and 2019 (n = 69) was compared with patients who did not participate in the programme, matched on sex, age, diabetes type and place of the general practitioner’s practice (n = 138). The primary outcome was glycated haemoglobin (HbA1c) change between 12 months before and 12 months after the DSM programme. Secondary outcomes included modifications in diabetes treatment, body mass index, blood pressure, dyslipidaemia, presence of microalbuminuria, and diabetes retinopathy screening participation.ResultsHbA1c was significantly improved in the exposed group after the programme (p < 0.01). The analysis did not find any significant between-group difference in socio-demographic data, medical history, comorbidities, and treatment adaptation.ConclusionsThese results, consistent with the international literature, promote the development of DSM programmes in primary care settings in deprived areas. The results of this real-life study need to be confirmed on the long-term and in different contexts (rural area, healthcare organisation).

Measuring Outcomes of Diabetic Retinopathy Screening: What Is Important?

Measuring Outcomes of Diabetic Retinopathy Screening: What Is Important?

Comparison of 21 artificial intelligence algorithms in automated diabetic retinopathy screening using handheld fundus camera

Background Diabetic retinopathy (DR) is a common complication of diabetes and may lead to irreversible visual loss. Efficient screening and improved treatment of both diabetes and DR have amended visual prognosis for DR. The number of patients with diabetes is increasing and telemedicine, mobile handheld devices and automated solutions may alleviate the burden for healthcare. We compared the performance of 21 artificial intelligence (AI) algorithms for referable DR screening in datasets taken by handheld Optomed Aurora fundus camera in a real-world setting. Patients and methods Prospective study of 156 patients (312 eyes) attending DR screening and follow-up. Both papilla- and macula-centred 50° fundus images were taken from each eye. DR was graded by experienced ophthalmologists and 21 AI algorithms. Results Most eyes, 183 out of 312 (58.7%), had no DR and mild NPDR was noted in 21 (6.7%) of the eyes. Moderate NPDR was detected in 66 (21.2%) of the eyes, severe NPDR in 1 (0.3%), and PDR in 41 (13.1%) composing a group of 34.6% of eyes with referable DR. The AI algorithms achieved a mean agreement of 79.4% for referable DR, but the results varied from 49.4% to 92.3%. The mean sensitivity for referable DR was 77.5% (95% CI 69.1–85.8) and specificity 80.6% (95% CI 72.1–89.2). The rate for images ungradable by AI varied from 0% to 28.2% (mean 1.9%). Nineteen out of 21 (90.5%) AI algorithms resulted in grading for DR at least in 98% of the images. Conclusions Fundus images captured with Optomed Aurora were suitable for DR screening. The performance of the AI algorithms varied considerably emphasizing the need for external validation of screening algorithms in real-world settings before their clinical application.

Deep learning-driven automated quality assessment of ultra-widefield optical coherence tomography angiography images for diabetic retinopathy

AbstractImage quality assessment (IQA) of fundus images constitutes a foundational step in automated disease analysis. This process is pivotal in supporting the automation of screening, diagnosis, follow-up, and related academic research for diabetic retinopathy (DR). This study introduced a deep learning-based approach for IQA of ultra-widefield optical coherence tomography angiography (UW-OCTA) images of patients with DR. Given the novelty of ultra-widefield technology, its limited prevalence, the high costs associated with equipment and operational training, and concerns regarding ethics and patient privacy, UW-OCTA datasets are notably scarce. To address this, we initially pre-train a vision transformer (ViT) model on a dataset comprising 6 mm × 6 mm OCTA images, enabling the model to acquire a fundamental understanding of OCTA image characteristics and quality indicators. Subsequent fine-tuning on 12 mm × 12 mm UW-OCTA images aims to enhance accuracy in quality assessment. This transfer learning strategy leverages the generic features learned during pre-training and adjusts the model to evaluate UW-OCTA image quality effectively. Experimental results demonstrate that our proposed method achieves superior performance compared to ResNet18, ResNet34, and ResNet50, with an AUC of 0.9026 and a Kappa value of 0.7310. Additionally, ablation studies, including the omission of pre-training on 6 mm × 6 mm OCTA images and the substitution of the backbone network with the ViT base version, resulted in varying degrees of decline in AUC and Kappa values, confirming the efficacy of each module within our methodology.

Cost-Effectiveness of Diabetic Retinopathy Screening for Newly Diagnosed Type 2 Diabetic Patients: A Nationwide Population-based Propensity Score-Matched Cohort Study

AimsThis study investigated the association between the frequency of screening for diabetic retinopathy (DR) versus the development of DR and corresponding medical expenses among patients newly diagnosed with type 2 diabetes mellitus (T2DM). MethodsThis longitudinal, population-based study used the Taiwan National Health Insurance Research Database (2004 to 2020) as a data source. Propensity score matching (PSM) (sex, age, comorbidities and concurrent medication use) was employed in the grouping of T2DM patients according to different frequency of DR screening. Outcome measures included the proportion of patients who developed DR, who received DR treatment, and the associated medical expenses and hospitalizations. ResultsThe 17-year cohort included 337,046 patients. After PSM, three groups each containing 35,739 patients were assembled and analyzed. Compared to low-frequency screening, high-frequency screening was more effective in detecting patients requiring treatment; however, the net cost for treatment was significantly lower. Standard-frequency screening appears to provide the best balance in terms of DR detection, diagnosis interval, the risk of DR-related hospitalization, and DR treatment costs. ConclusionsIn this real-world cohort study covering all levels of the healthcare system, infrequent screening was associated with delayed diagnosis and elevated treatment costs, while a fundus screening interval of 1-2 years proved optimal in terms of detection and medical expenditures.

Transition probabilities of diabetic retinopathy and death in an Asian population with Diabetes

PurposeTo evaluate the dynamic transitions in diabetic retinopathy (DR) severity over time and associated risk factors in an Asian population with diabetes. DesignLongitudinal cohort study MethodsWe analyzed data from 9481 adults in the Singapore Integrated Diabetic Retinopathy Screening Program (2010–2015) with linkage to death registry. A multistate Markov model adjusted for age, sex, systolic blood pressure (SBP), diabetes duration, HbA1c, and body mass index (BMI) was applied to estimate annual transition probabilities between four DR states (no, mild, moderate, and severe/proliferative) and death, and the mean sojourn time in each state. ResultsThe median assessment interval was 12 months, with most patients having 3 assessments. Annual probabilities for DR progression (no-to-mild, mild-to-moderate and moderate-to-severe/proliferative) were 6.1%, 7.0% and 19.3%, respectively; and for regression (mild-to-no, moderate-to-mild and severe-to-moderate) were 55.4%, 17.3% and 4.4%, respectively. Annual mortality rates from each DR state were 1.2%, 2.0%, 18.7%, and 30.0%. The sojourn time in each state were 8.2, 0.8, 0.8 and 2.2 years. Higher HbA1c and SBP levels were associated with progression of no-mild and mild-moderate DR, and diabetes duration with no-to-mild and moderate-to-severe/proliferative DR. Lower HbA1c levels were associated with regression from mild-to-no and moderate-to-mild, and higher BMI with mild-to-no DR. ConclusionsOur results suggest a prolonged duration (~8 years) in developing mild DR, with faster transitions (within a year) from mild or moderate states. Moderate/above DR greatly increases the probability of progression and death as compared to mild DR/below. HbA1c was associated with both progression as well as regression.

Synchronous Diagnosis of Diabetic Retinopathy by a Handheld Retinal Camera, Artificial Intelligence, and Simultaneous Specialist Confirmation

PurposeDiabetic retinopathy (DR) is a leading cause of preventable blindness, particularly in underserved regions where access to ophthalmic care is limited. This study presents a proof of concept for utilizing a portable handheld retinal camera with an embedded artificial intelligence (AI) platform, complemented by a synchronous remote confirmation by retina specialists, for DR screening in an underserved rural area. DesignRetrospective cohort study Subjects1,115 individuals with diabetes. MethodsA retrospective analysis of a screening initiative conducted in four municipalities in Northeastern Brazil, targeting the diabetic population. A portable handheld retinal camera captured macula-centered and disc-centered images, which were analyzed by the AI system. Immediate push notifications were sent out to retina specialists upon the detection of significant abnormalities, enabling synchronous verification and confirmation, with on-site patient feedback within minutes. Referral criteria were established, and all referred patients underwent a complete ophthalmic work-up and subsequent treatment. Main outcome measuresProof-of-concept implementation success. ResultsOut of 2,052 invited individuals, 1,115 participated, with a mean age of 60.93 years and diabetes duration of 7.52 years, of whom 66.03% were women. The screening covered 2,222 eyes, revealing various retinal conditions. Referable eyes for DR were 11.84%, with an additional 13% for other conditions (diagnoses included various stages of DR, media opacity, nevus, drusen, enlarged cup-to-disc ratio, pigmentary changes, and other). AI performance for overall detection of referable cases (both DR and other conditions) was: sensitivity- 84.23% (95% CI- 82.63-85.84), specificity- 80.79% (95% CI- 79.05- 82.53). When we assessed whether AI matched any clinical diagnosis, be it referable or not, sensitivity was 85.67% (95% CI- 84.12-87.22), specificity was 98.86 (95% CI- 98.39-99.33), and AUC was 0.92 (95% CI- 0.91-0.94). ConclusionsThe integration of a portable device, AI analysis, and synchronous medical validation has the potential to play a crucial role in preventing blindness from DR, especially in socially unequal scenarios.

Comparing the Implementation of Different Diabetic Retinopathy Screening Models in Primary Health Care Settings in Northern India: Pragmatic Three-Arm Observational Study

Introduction: Early screening is recommended to prevent vision impairment (VI) and blindness globally due to diabetic retinopathy (DR). We aim to analyze the factors influencing the implementation of DR screening models in primary healthcare facilities. Methods: A pragmatic three-arm observational study will be conducted in the district of Mohali in Punjab, India. The study will compare three equal-sized groups with type 2 diabetes mellitus (T2DM) at age 30 and above. The T2DM participants and healthcare providers (HCPs) will be interviewed to identify the barriers to retinal examinations in primary healthcare settings. The study includes a comprehensive economic evaluation to determine the cost-effectiveness of implementing different DR screening models in primary health settings. The present study will use the RE-AIM framework (reach, effectiveness, adoption, implementation, and maintenance), to assess the implementation of two DR screening interventions and the extent of variation in outcomes. Qualitative and quantitative data collected from participants and research team observations will be examined to report on the five RE-AIM domains. Discussion: A screening program should ensure early detection and referral for DR treatment before it results in irreversible vision loss. However, understanding the problems of accessing DR screening and estimating the cost-effectiveness of such programs in a primary health setting is important before allocating public health resources.

Classification of Diabetic Retinopathy using Deep Learning

Abstract: Diabetic retinopathy, a potentially fatal retinal disease, occurs in individuals with diabetes, causing vision damage and possible visual impairment traditionally, screening for diabetic retinopathy has been the norm hard by ophthalmologists manually. To facilitate this, we have DR. Our pre-trained model is trained on a large dataset of about 3662 training images, enabling it to automatically detect the DR platform in high-resolution fundus images The dataset used for this is publicly available on kaggle. The DR phases are categorized as 0, 1, 2, 3, and 4. The fundus eye images serve as the input parameters of the model. The pre-trained model then identifies the point of interest in bank eye images, then provides activation function insights By calculating weights, for example patients’ visual intensity levels, it helps classify diabetic retinopathy images into their intensity groups appropriately within the possibility of test- methods Provides variation.

Patients with Type 2 Diabetes, Higher Blood Pressure, and Infrequent Fundus Examinations Have a Higher Risk of Sight-Threatening Retinopathy.

Background: Diabetic retinopathy (DR) is the most common cause of preventable blindness among working-age adults. This study aimed to evaluate the impact of the regularity of fundus examinations and risk factor control in patients with type 2 diabetes (T2DM) on the prevalence and severity of DR. Methods: One hundred and fifty-six T2DM patients were included in this cross-sectional study. Results: In this sample, the prevalence of DR was 46.2%. Patients with no DR mainly did not examine the fundus regularly, while most patients with mild/moderate nonproliferative DR (NPDR) underwent a fundus examination regularly. In 39.7% of patients, this was the first fundus examination due to diabetes, and 67% of them had sight-threatening DR (STDR). Diabetes duration (p = 0.007), poor glycemic control (HbA1c) (p = 0.006), higher systolic blood pressure (SBP) (p < 0.001), and diastolic blood pressure (DBP) (p = 0.002) were the main predictors of DR. However, the impact of SBP (AOR 1.07, p = 0.003) and DBP (AOR 1.13, p = 0.005) on DR development remained significant even after adjustment for diabetes duration and HbA1c. The DR prevalence was higher in patients with higher blood pressure (≥130/80 mmHg) than in those with target blood pressure (<130/80 mmHg) (p = 0.043). None of the patients with target blood pressure had STDR. The peaks in SBP and DBP were observed in T2DM with DR and the first fundus examination due to diabetes. Conclusions: In this T2DM sample, DR prevalence was very high and strongly related to blood pressure and a lack of regular fundus examinations. These results indicate the necessity of establishing systematic DR screening in routine diabetes care and targeting blood pressure levels according to T2DM guidelines.

Autonomous artificial intelligence versus teleophthalmology for diabetic retinopathy.

Purpose: To assess the role of artificial intelligence (AI) based automated software for detection of Diabetic Retinopathy (DR) compared with the evaluation of digital retinography by two double masked retina specialists. Methods: Two-hundred one patients (mean age 65 ± 13 years) with type 1 diabetes mellitus or type 2 diabetes mellitus were included. All patients were undergoing a retinography and spectral domain optical coherence tomography (SD-OCT, DRI 3D OCT-2000, Topcon) of the macula. The retinal photographs were graded using two validated AI DR screening software (Eye Art TM and IDx-DR) designed to identify more than mild DR. Results: Retinal images of 201 patients were graded. DR (more than mild DR) was detected by the ophthalmologists in 38 (18.9%) patients and by the AI-algorithms in 36 patients (with 30 eyes diagnosed by both algorithms). Ungradable patients by the AI software were 13 (6.5%) and 16 (8%) for the Eye Art and IDx-DR, respectively. Both AI software strategies showed a high sensitivity and specificity for detecting any more than mild DR without showing any statistically significant difference between them. Conclusions: The comparison between the diagnosis provided by artificial intelligence based automated software and the reference clinical diagnosis showed that they can work at a level of sensitivity that is similar to that achieved by experts.

Referrals for proliferative diabetic retinopathy from two UK diabetic retinopathy screening services: a 10-year analysis of visual outcomes, requirement for vitrectomy, and mortality.

To determine long-term outcomes of patients referred with proliferative diabetic retinopathy (PDR) from diabetic eye screening programmes (DESP) to tertiary care centres in the United Kingdom (UK). Retrospective multicentre study of patients referred from two DESPs in the UK over a 36-month period (2007-9) and followed-up for 10 years. Critical outcomes included severe vision loss (SVL) and the need for vitrectomy. Other outcomes assessed included moderate vision loss (MVL), and patient survival time. Univariate and multiple variable Cox proportional hazards regressions were used to analyse survival outcomes. 212 eyes of 150 patients were referred with a diagnosis of PDR. 109 eyes of 72 patients were confirmed to have active PDR and included in the study. 61% of patients had low-risk PDR, while 39% exhibited high-risk features in at least one eye. Eight (7.3%) eyes developed SVL and 16 (14.7%) MVL during follow up. Vitrectomy was required in 24% (95% CI: 15 to 31%) of all PDR eyes and was most commonly performed for vitreous haemorrhage (65%). The 10-year survival in all PDR patients was 76% (95% CI: 63 to 85%) with the mean time to death for all deceased patients being 5.4 ± 3.6 years. On multivariable analysis, only age was found to have a significant association with the survival of patients with PDR. During the 10 year follow up SVL was uncommon, but MVL occurred in almost one-fifth of the eyes. Approximately 1 in 4 eyes required vitrectomy, highlighting its significance in patient management.

Referable Diabetic Retinopathy Prediction Algorithm Applied to a Population of 120,389 Type 2 Diabetics over 11 Years Follow-Up.

(1) Background: Although DR screening is effective, one of its most significant problems is a lack of attendance. The aim of the present study was to demonstrate the effectiveness of our algorithm in predicting the development of any type of DR and referable DR. (2) Methods: A retrospective study with an 11-year follow-up of a population of 120,389 T2DM patients was undertaken. (3) Results: Applying the results of the algorithm showed an AUC of 0.93 (95% CI, 0.92-0.94) for any DR and 0.90 (95% CI, 0.89-0.91) for referable DR. Therefore, we achieved a promising level of agreement when applying our algorithm. (4) Conclusions: The algorithm is useful for predicting which patients may develop referable forms of DR and also any type of DR. This would allow a personalized screening plan to be drawn up for each patient.

RLeU-Net: Segmentation of blood vessels in retinal fundus images for Diabetic Retinopathy Screening

RLeU-Net: Segmentation of blood vessels in retinal fundus images for Diabetic Retinopathy Screening

Assessment of prevalence and need for screening of diabetic retinopathy using non-mydriatic fundus camera in rural and tribal diabetic populations in Maharashtra.

To assess the prevalence of DR and the need for screening and management of DR with medical management of diabetes in rural and tribal population in Maharashtra. The known diabetics of rural area and tribal area were screened at corresponding primary health centers, subcenters, and village level with the help of local healthcare workers using a portable non-mydriatic fundus camera. The prevalence of blindness among known diabetics in rural area was 1.29%, and 0.84% in tribal area. In the rural area, the prevalence of diabetic retinopathy (DR) was 5.67% (n = 776), out of those 18.18% had sight threatening diabetic retinopathy (STDR). The prevalence of DR was 7.73% (n = 711) in tribal areas, out of those, 30.90% had STDR. The significant risk factors were identified to be the duration of diabetes and poor glycemic control. Implementation of targeted interventions for screening and management are required to reduce the risk of blindness among known diabetics in rural and tribal areas.

Novel Technologies in Artificial Intelligence and Telemedicine for Glaucoma Screening.

Please note that this content has temporarily been taken down as it was inadvertently posted before it was fully ready for publication.

Existing and emerging GLP-1 receptor agonist therapy: Ramifications for diabetic retinopathy screening.

GLP-1 receptor agonist treatment revolutionised the management of type 2 diabetes mellitus with significant enhancement of cardiovascular risk reduction. They have been instrumental in effectively managing the glycaemic control of this at-risk patient group. This class of drugs are associated with rapid improvement in glucose levels and consequently, transient early worsening of pre-existing diabetic retinopathy (DR) which is well-recognised, but this paradox is less commonly perceived in routine clinical practice. The recent shortage of supply has resulted in an enforced hiatus to prescribing all existing GLP-1 receptor agonists, which is expected to last all through 2024. This becomes even more pertinent as their DR could have progressed due to worsening HbA1c as a result of the unforeseen interruption to GLP-1 receptor agonist treatment. Therefore, when these medications are recommenced in a few months' time, all prescribers need to be aware of these patients' most up-to-date DR status and liaise with their affiliated screening service.

Attention-based deep learning framework for automatic fundus image processing to aid in diabetic retinopathy grading

Background and objectiveEarly detection and grading of Diabetic Retinopathy (DR) is essential to determine an adequate treatment and prevent severe vision loss. However, the manual analysis of fundus images is time consuming and DR screening programs are challenged by the availability of human graders. Current automatic approaches for DR grading attempt the joint detection of all signs at the same time. However, the classification can be optimized if red lesions and bright lesions are independently processed since the task gets divided and simplified. Furthermore, clinicians would greatly benefit from explainable artificial intelligence (XAI) to support the automatic model predictions, especially when the type of lesion is specified. As a novelty, we propose an end-to-end deep learning framework for automatic DR grading (5 severity degrees) based on separating the attention of the dark structures from the bright structures of the retina. As the main contribution, this approach allowed us to generate independent interpretable attention maps for red lesions, such as microaneurysms and hemorrhages, and bright lesions, such as hard exudates, while using image-level labels only. MethodsOur approach is based on a novel attention mechanism which focuses separately on the dark and the bright structures of the retina by performing a previous image decomposition. This mechanism can be seen as a XAI approach which generates independent attention maps for red lesions and bright lesions. The framework includes an image quality assessment stage and deep learning-related techniques, such as data augmentation, transfer learning and fine-tuning. We used the architecture Xception as a feature extractor and the focal loss function to deal with data imbalance. ResultsThe Kaggle DR detection dataset was used for method development and validation. The proposed approach achieved 83.7 % accuracy and a Quadratic Weighted Kappa of 0.78 to classify DR among 5 severity degrees, which outperforms several state-of-the-art approaches. Nevertheless, the main result of this work is the generated attention maps, which reveal the pathological regions on the image distinguishing the red lesions and the bright lesions. These maps provide explainability to the model predictions. ConclusionsOur results suggest that our framework is effective to automatically grade DR. The separate attention approach has proven useful for optimizing the classification. On top of that, the obtained attention maps facilitate visual interpretation for clinicians. Therefore, the proposed method could be a diagnostic aid for the early detection and grading of DR.

Effectiveness of Conventional Digital Fundus Photography-Based Teleretinal Screening for Diabetic Retinopathy and Diabetic Macular Edema: A Report by the American Academy of Ophthalmology

This American Academy of Ophthalmology Ophthalmic Technology Assessment aims to assess the effectiveness of conventional teleretinal screening (TS) in detecting diabetic retinopathy (DR) and diabetic macular edema (DME). A literature search of the PubMed database was conducted most recently in July 2023 to identify data published between 2006 and 2023 on any of the following elements related to TS effectiveness: (1) the accuracy of TS in detecting DR or DME compared with traditional ophthalmic screening with dilated fundus examination or 7-standard field Early Treatment Diabetic Retinopathy Study photography, (2) the impact of TS on DR screening compliance rates or other patient behaviors, and (3) cost-effectiveness and patient satisfaction of TS compared with traditional DR screening. Identified studies then were rated based on the Oxford Centre for Evidence-Based Medicine grading system. Eight level I studies, 14 level II studies, and 2 level III studies were identified in total. Although cross-study comparison is challenging because of differences in reference standards and grading methods, TS demonstrated acceptable sensitivity and good specificity in detecting DR; moderate to good agreement between TS and reference-standard DR grading was observed. Performance of TS was not as robust in detecting DME, although the number of studies evaluating DME specifically was limited. Two level I studies, 5 level II studies, and 1 level III study supported that TS had a positive impact on overall DR screening compliance, even increasing it by more than 2-fold in one study. Studies assessing cost-effectiveness and patient satisfaction were not graded formally, but they generally showed that TS was cost-effective and preferred by patients over traditional surveillance. Conventional TS is an effective approach to DR screening not only for its accuracy in detecting referable-level disease, but also for improving screening compliance in a cost-effective manner that may be preferred by patients. Further research is needed to elucidate the ideal approach of TS that may involve integration of artificial intelligence or other imaging technologies in the future. Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Compound Scaling Encoder-Decoder (CoSED) Network for Diabetic Retinopathy Related Bio-marker Detection.

Biomedical image segmentation plays an important role in Diabetic Retinopathy (DR)-related biomarker detection. DR is an ocular disease that affects the retina in people with diabetes and could lead to visual impairment if management measures are not taken in a timely manner. In DR screening programs, the presence and severity of DR are identified and classified based on various microvascular lesions detected by qualified ophthalmic screeners. Such a detection process is time-consuming and error-prone, given the small size of the microvascular lesions and the volume of images, especially with the increasing prevalence of diabetes. Automated image processing using deep learning methods is recognized as a promising approach to support diabetic retinopathy screening. In this paper, we propose a novel compound scaling encoder-decoder network architecture to improve the accuracy and running efficiency of microvascular lesion segmentation. In the encoder phase, we develop a lightweight encoder to speed up the training process, where the encoder network is scaled up in depth, width, and resolution dimensions. In the decoder phase, an attention mechanism is introduced to yield higher accuracy. Specifically, we employ Concurrent Spatial and Channel Squeeze and Channel Excitation (scSE) blocks to fully utilise both spatial and channel-wise information. Additionally, a compound loss function is incorporated with transfer learning to handle the problem of imbalanced data and further improve performance. To assess performance, our method is evaluated on two large-scale lesion segmentation datasets: DDR and FGADR datasets. Experimental results demonstrate the superiority of our method compared to other competent methods. Our codes are available at https://github.com/DeweiYi/CoSED-Net.