Longitudinal Quantitative Ultrawidefield Angiographic Features in Diabetic Retinopathy Treated with Aflibercept from the Intravitreal Aflibercept as Indicated by Real-Time Objective Imaging to Achieve Diabetic Retinopathy Improvement Trial

OBJECTIVE

To report longitudinal trends of quantitative ultrawidefield fluorescein angiography (qUWFA) biomarkers in the Intravitreal Aflibercept as Indicated by Real-Time Objective Imaging to Achieve Diabetic Retinopathy Improvement (PRIME) diabetic retinopathy (DR) clinical trial.

DESIGN

Post hoc analysis of the PRIME prospective randomized DR clinical trial comparing intravitreal aflibercept treatment based on the DR severity score (DRSS) or quantitative leakage index for DR improvement (ClinicalTrials.gov identifier: NCT03531294).

PARTICIPANTS

Patients were enrolled with a DRSS level of 47A to 71A and best-corrected visual acuity of 20/800 or better. Key exclusion criteria were previous intravitreal injection, panretinal photocoagulation, vitrectomy, central-involving macular edema, or vitreous hemorrhage.

METHODS

A previously validated, machine learning-based qUWFA analysis platform was used for panretinal leakage index assessment and differentiation of generalized and perivascular leakage phenotypes. Additionally, microaneurysm count and ischemic index were quantified in panretinal and macular regions. The trends in these biomarkers and therapeutic response were studied over 1 year.

MAIN OUTCOME MEASURES

Longitudinal trends of qUWFA biomarkers. The impact of these qUWFA metrics on treatment response was assessed by studying their associations with time to 2-step DRSS improvement and number of treatment-free days.

RESULTS

Forty eyes from 40 subjects with DR were enrolled. Lower baseline generalized leakage was noted in eyes that attained the 2-step DRSS improvement in < 16 weeks (1.9% vs. 2.8%; P = 0.026). Baseline macular perivascular-generalized leakage ratio had a significant correlation with the number of treatment-free days (r = 0.4; P = 0.012). At the end of 1 year, therapy significantly reduced the mean panretinal (3.9% vs. 5.8%; P = 0.002) and macular (6.2% vs. 12.2%; P = 0.008) generalized leakage indices compared with baseline, as well as the mean panretinal perivascular leakage index (1.5% vs. 2.3%; P = 0.002). The mean panretinal ischemic index demonstrated a small but likely clinically insignificant decrease from 12.5% at baseline to 11.6% at year 1 (P = 0.016).

CONCLUSIONS

Down-trending leakage indices and microaneurysm counts were demonstrated over 1 year of anti-VEGF therapy. At baseline, DR eyes with lower generalized leakage responded to therapy more rapidly. Eyes with greater perivascular leakage relative to generalized leakage showed a longer-lasting anti-VEGF treatment response.

FINANCIAL DISCLOSURE(S)

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

An optimized framework for prediagnosis of diabetic retinopathy using HHO-CBL model

BACKGROUND

Diabetic Retinopathy (DR) is a widespread intense stage of diabetes mellitus that causes vision-effecting anomalies in the retina. It is a medical health condition on the strength of fluctuating glucose level in the blood that can result in vision loss in case of severity.

OBJECTIVE

As a result, early detection and treatment with DR is the most significant task which will tremendously reduce the likelihood of vision impairment and is still a difficult challenge. Many conventional methods fail to detect primary causes of formation of Microaneurysms, that are used to determine the Prediagnosis of DR.

METHOD

To overcome this challenge, the proposed model incorporates Harris Hawk Optimization with CNN-Bi-LSTM (HHO-CBL) to extract the features. The Prediagnosis of DR has been achieved through this model by spotting saccular dilations, hyaline like material in the capillary aneurysm wall, kinking of vessels since these are the indications for the creation of microaneurysms that are spotted in the blood vessel of the retina. The recommended model is also used to automatically detect DR and its progression in many phases. Furthermore, in order to identify the severity of DR retina, we used a benchmark Kaggle APTOS dataset to train the HHO-CBL model.

RESULTS

Experimental results reveal that this model obtains the best classification accuracy of 96.4 % for an early diagnosis and 98.8 % for a five-degree classification. In addition to those results, a comparison with previously carried out studies has also shown that this model provides a promising solution for a successful Prediagnosis of DR and its staging.

CONCLUSIONS

In the current research, an innovative HHO-CBL was developed for identifying the primary causes that lead to the formation of microaneurysms and diagnosing all five grades of DR. According to the acquired results presented through the evaluation performance metrics indicates that the pre-early diagnosis and five grade classification using feature embedding technique outperformed the other prevailing approaches (Tab. 4, Fig. 10, Ref. 31).

RETINAL HYPOPERFUSION DETECTED BY WIDE-FIELD OPTICAL COHERENCE TOMOGRAPHIC ANGIOGRAPHY IN A CASE OF RETINAL RACEMOSE HEMANGIOMA

PURPOSE

Retinal racemose hemangioma is a rare, unilateral, nonhereditary, arteriovenous malformation characterized by the appearance of dilated and tortuous retinal vessels. Retinal racemose hemangioma can develop complications associated with retinal ischemia, such as vitreous hemorrhage, retinal vein occlusion, and neovascular glaucoma. Here, a case of retinal racemose hemangioma with retinal hypoperfusion detected by wide-field swept-source optical coherence tomographic angiography was reported, which was not unambiguously illustrated by fluorescein angiography.

METHODS

Case report.

RESULTS

A 57-year-old woman was referred to our hospital for the evaluation of severe retinal vascular tortuosity, dilation, and retinal hemorrhages in the left eye. Fundus examination revealed arteriovenous communications temporal to the fovea and multiple microaneurysms surrounded by retinal hemorrhages at the midperipheral temporal fundus. In fluorescein angiography, multiple hyperfluorescent lesions with leakage corresponding to microaneurysms were observed in the temporal and lower midperipheral areas; however, nonperfused areas were apparently absent. By contrast, wide-field optical coherence tomographic angiography clearly showed low-density retinal capillaries in the superotemporal quadrant in comparison with those in the inferotemporal quadrant.

CONCLUSION

Wide-field optical coherence tomographic angiography detected sparse retinal capillaries, which were not well illustrated by fluorescein angiography, in a patient with retinal racemose hemangioma. This indicates the presence of low-grade retinal hypoperfusion caused by altered retinal hemodynamics, potentially leading to ischemia-related retinal disorders during a prolonged course, in patients with clinically quiescent retinal racemose hemangioma.

Retinopathy is associated with impaired cognition in patients undergoing peritoneal dialysis

Objective: Previous studies have shown a relationship between retinopathy and cognition including population with and without chronic kidney disease (CKD) but data regarding peritoneal dialysis (PD) are limited. This study aims to investigate the relationship between retinopathy and cognitive impairment in patients undergoing peritoneal dialysis (PD). Methods: In this observational study, we recruited a total of 107 participants undergoing PD, consisting of 48 men and 59 women, ages ranging from 21 to 78 years. The study followed a cross-sectional design. Retinal microvascular characteristics, such as geometric changes in retinal vascular including tortuosity, fractal dimension (FD), and calibers, were assessed. Retinopathy (such as retinal hemorrhage or microaneurysms) was evaluated using digitized photographs. The Modified Mini-Mental State Examination (3MS) was performed to assess global cognitive function. Results: The prevalence rates of retinal hemorrhage, microaneurysms, and retinopathy were 25%, 30%, and 43%, respectively. The mean arteriolar and venular calibers were 63.2 and 78.5 µm, respectively, and the corresponding mean tortuosity was 37.7 ± 3.6 and 37.2 ± 3.0 mm-1. The mean FD was 1.49. After adjusting for age, sex, education, mean arterial pressure, and Charlson index, a negative association was revealed between retinopathy and 3MS scores (regression coefficient: -3.71, 95% confidence interval: -7.09 to -0.33, p = 0.03). Conclusions: Retinopathy, a condition common in patients undergoing PD, was associated with global cognitive impairment. These findings highlight retinopathy, can serve as a valuable primary screening tool for assessing the risk of cognitive decline.

New retinal findings in NLRP3-associated autoinflammatory disease

PURPOSE

To determine whether the rare NLRP3-Associated Autoinflammatory Disease (NLRP3-AID) is associated with retinal changes and to assess the ocular involvement.

METHODS

A retrospective cohort study of 20 patients(40 eyes) diagnosed with rare NLRP3-AID at Peking Union Medical College Hospital, from April 2015 to August 2022. Patients underwent a comprehensive ophthalmological examination, including visual acuity, intraocular pressure examination, slit-lamp examination, fundus photography, optical coherence tomography(OCT), and fluorescence angiography (FA). Some patients also underwent optical coherence tomography angiography (OCTA).

RESULTS

This study analyzed 40 eyes of 20 patients (11 [55.0%] male; median age, 25.0 years [range, 12-52 years]) and 13 patients (26 eyes, 65%) demonstrated ocular involvement. The most common ophthalmologic manifestation was conjunctivitis (22 eyes, 84.6%), followed by papilledema (14 eyes, 53.8%), retinopathy (10 eyes, 38.5%), optic atrophy (6 eyes, 23.1%), uveitis (4 eyes, 15.4%), reduced pupil light reflex (3 eyes, 11.5%) and cataracts (2 eyes, 7.7%). Ocular involvement was bilateral in 11 patients (55.0%). Five kinds of retinal lesions were seen in 5 patients (10 eyes, 25%) with NLRP3-AID, including peripheral retinal vascular leakage, microaneurysms, macular ischemia, macular epiretinal membrane formation and drusen.

CONCLUSIONS

Peripheral retinal vascular leakage, macular ischemia, microaneurysms and drusen are newly identified retinal findings in patients with NLRP3-AID, which suggests the importance of detailed retinal examination in these patients.

UK Biobank retinal imaging grading: methodology, baseline characteristics and findings for common ocular diseases

BACKGROUND/OBJECTIVES

This study aims to describe the grading methods and baseline characteristics for UK Biobank (UKBB) participants who underwent retinal imaging in 2009-2010, and to characterise individuals with retinal features suggestive of age-related macular degeneration (AMD), glaucoma and retinopathy.

METHODS

Non-mydriatic colour fundus photographs and macular optical coherence tomography (OCT) scans were manually graded by Central Administrative Research Facility certified graders and quality assured by clinicians of the Network of Ophthalmic Reading Centres UK. Captured retinal features included those associated with AMD (≥1 drusen, pigmentary changes, geographic atrophy or exudative AMD; either imaging modality), glaucoma (≥0.7 cup-disc ratio, ≥0.2 cup-disc ratio difference between eyes, other abnormal disc features; photographs only) and retinopathy (characteristic features of diabetic retinopathy with or without microaneurysms; either imaging modality). Suspected cases of these conditions were characterised with reference to diagnostic records, physical and biochemical measurements.

RESULTS

Among 68,514 UKBB participants who underwent retinal imaging, the mean age was 57.3 years (standard deviation 8.2), 45.7% were men and 90.6% were of White ethnicity. A total of 64,367 participants had gradable colour fundus photographs and 68,281 had gradable OCT scans in at least one eye. Retinal features suggestive of AMD and glaucoma were identified in 15,176 and 2184 participants, of whom 125 (0.8%) and 188 (8.6%), respectively, had a recorded diagnosis. Of 264 participants identified to have retinopathy with microaneurysms, 251 (95.1%) had either diabetes or hypertension.

CONCLUSIONS

This dataset represents a valuable addition to what is currently available in UKBB, providing important insights to both ocular and systemic health.

THE LONG-TERM EFFECT OF CYSTOTOMY WITH FIBRINOGEN CLOT REMOVAL FOR A CYSTOID MACULAR EDEMA SECONDARY TO IDIOPATHIC MACULAR TELANGIECTASIA TYPE 1: A CASE REPORT

PURPOSE

To report the case of a patient with cystoid macular edema secondary to idiopathic macular telangiectasia (MacTel) Type 1, which was successfully treated by cystotomy and en bloc removal of the fibrinogen-rich component of the cystoid lesion.

METHODS

An 80-year-old man was referred to our department because of a visual defect in his right eye. His best-corrected decimal visual acuity was 0.7 (Snellen equivalent, 20/30). A fundus examination revealed clustered temporal juxafoveal microaneurysms and foveal cystoid macular edema. The patient refused to undergo conventional treatments, including direct retinal photocoagulation for microaneurysms, intravitreal anti-vascular endothelial growth factor injection, and intravitreal triamcinolone injection. However, he provided consent to undergo cystotomy and en bloc removal of the fibrinogen-rich component of the cystoid lesion.

RESULTS

His best-corrected decimal visual acuity was 0.2 (Snellen equivalent, 20/100) just before the surgery. A 27-gauge vitrectomy with internal limiting membrane peeling was performed. Cystotomy was performed during the surgery, and the fibrinogen clot visible in the cystoid cavity was also removed. Cystoid macular edema rapidly disappeared after the surgery. Three years postoperatively, the patient had best-corrected decimal visual acuity of 0.5 (Snellen equivalent, 20/40) at the last medical examination, and the cystoid macular edema had not recurred.

CONCLUSION

Cystotomy and en bloc removal of the fibrinogen-rich component of the cystoid lesion could be valid treatment options for cystoid macular edema secondary to MacTel Type 1.

Exploring the angiographic-biologic phenotype in the IMAGINE study: quantitative UWFA and cytokine expression

BACKGROUND

This study investigates the association of intraocular cytokine expression and ultrawide-field fluorescein angiography (UWFA) quantitative imaging biomarkers and their association with angiographical feature response after antivascular endothelial growth factor (VEGF) therapy in diabetic macular oedema (DME).

METHODS

The IMAGINE DME study is a post hoc imaging biomarker and intraocular cytokine assessment from the DAVE study, a prospective DME clinical trial that included aqueous humour sampling and UWFA imaging. Fifty-four cytokines associated with inflammation and angiogenesis were evaluated through multiplex arrays. UWFA parameters were assessed using an automated feature analysis platform to determine ischaemic and leakage indices and microaneurysm (MA) count. Eyes were classified into UWFA responder or non-responder groups based on longitudinal quantitative UWFA parameter improvement. Cytokine expression was correlated with UWFA metrics and evaluated in the context of therapeutic response.

RESULTS

Twenty-one eyes were included with a mean age of 55±10 years. Increased panretinal leakage index correlated with VEGF (r=0.70, p=0.0005), angiopoietin-like 4 (r=0.77, p=4.6E-5) and interleukin (IL)-6 (r=0.64, p=0.002). Panretinal ischaemic index was associated with tissue inhibitor of metalloproteinases 1 (TIMP-1, r=0.49, p=0.03) and peripheral ischaemia correlated with VEGF (r=0.45, p=0.05). MA count correlated with increased monocyte chemotactic protein-4 (MCP-4, r=0.60, p=0.004) and platelet and endothelial cell adhesion molecule 1 (PECAM-1, r=0.58, p=0.005). Longitudinal MA reduction was associated with decreased baseline VEGF and urokinase receptor (uPAR) (p<0.05). High baseline VEGF and IL-6 were associated with dramatic reduction in macular leakage (p<0.05).

CONCLUSIONS

Baseline and longitudinal quantitative UWFA imaging parameters correlated with multiple aqueous humour cytokine concentrations, including VEGF and IL-6. Further research is needed to assess the possible implications of using these findings for evaluating treatment response.

Expression of Osteopontin and Integrin αvβ3 Receptor in Retina of Diabetic Guinea Pigs with High Myopia

INTRODUCTION

The study aimed to investigate the effect of high myopia on the expression of retinal osteopontin (OPN) and integrin αvβ3 receptor in guinea pigs and determine the relationship between high myopia and diabetic retinopathy (DR).

METHODS

Ninety-three-week-old male guinea pigs were randomly divided into four groups that included normal control group (NOR, n = 18), high myopia group (HM, n = 24), diabetes group (DR, n = 24), and diabetes with high myopia group (DR+HM, n = 24). HM was induced by form deprivation (FDHM) in the right eye. The DR group was injected with 5% streptozotocin 280 mg/kg intraperitoneally in the lower left abdomen of guinea pigs. The DRHM group was subjected to the same treatment as the HM and DR groups. Eighteen guinea pigs in each group were randomly selected to complete the experimental measurement. After enucleation of eyeballs, HE and immunohistochemical staining were performed to observe the retina morphology and count the positive rate of OPN and integrin αvβ3 receptor.

RESULTS

Diabetic retinal changes were found in group DR and HM+DR. The degree of retinal change in group HM+DR was less than that in group DR. In the DR group, the morphology of retinal tissue was loose, the number of cells decreased, retinal microaneurysms were increased, and a small amount of small artery embolism and venous thrombosis were observed. Although the retinal structure in the HM+DR group also became thinner, looser, and disordered, only a small number of microaneurysms were observed compared with the diabetic group. Immunohistochemical staining showed that the expression of OPN and integrin αvβ3 receptors in the diabetic groups (DR, HM+DR) was significantly higher than in the HM and NOR groups. The positive expression rates of OPN and integrin αvβ3 receptors in group HM+DR were significantly lower than those in group DR (p < 0.05).

CONCLUSION

The expression of OPN and integrin αvβ3 receptor in the retina of diabetic guinea pigs with high myopia was lower than that of diabetic models, which may be due to the influence of high myopia on neovascularization in DR.

Effect of a Fixed Combination of Curcumin, Artemisia, Bromelain, and Black Pepper Oral Administration on Optical Coherence Tomography Angiography Indices in Patients with Diabetic Macular Edema

Background: To investigate the effects of a fixed combination of Curcumin (200 mg), Artemisia (80 mg), Bromelain (80 mg), and Black pepper (2 mg) on vascular parameters in mild to moderate diabetic macular edema (DME). Design: Prospective, case-control study. Methods: Fifty-six patients affected by diabetes mellitus type II were enrolled in the study. Twenty-eight patients with DME received 2 tablets/day, before meals of a dietary complementary supplement containing in fixed combination Curcumin (200 mg), Artemisia (80 mg), Bromelain (80 mg), and Black pepper (2 mg) (Intravit®, OFFHEALTH Spa, Firenze, Italy) for 6 months. Twenty-eight age-matched subjects affected by diabetes mellitus type II were given placebo and served as control group. Patients underwent best correct visual acuity (BCVA), swept optical coherence tomography (OCT), and OCT-Angiography (OCTA). OCTA images of the superficial capillary plexus (SCP) and deep capillary plexus (DCP) were obtained for each eye. By the end of the follow-up patients were defined responder to the therapy when a decrease of more than 30 μm was registered in central retinal thickness (CRT) measurement, while a poor responder was determined by the absence of reduction or an increase in central retinal thickness at 6 months. We assessed the foveal avascular zone (FAZ) area, vessel density and quantified the number of microaneurysms in each layer. Results: A significant improvement of BCVA and CRT reduction was recorded at 6 months follow-up in the dietary complementary supplementation group compared to control (respectively p = 0.028 and p = 0.0003). VD of the total capillary plexus, microaneurysms count, glycaemia and HbA1c did not vary over the follow-up period between groups. Within the Intravit® group, poor responders tended to show a larger FAZ area, more microaneurysms, and a lower VD in the DCP compared to the good responders group (p < 0.0001). Conclusions: A fixed combination of Curcumin, Artemisia, Bromelain, and Black pepper oral administration may have a positive impact on central retinal thickness, visual acuity, and VD of the DCP in compensated type 2 diabetic patients with mild DME.

Detection of Fundus Lesions through a Convolutional Neural Network in Patients with Diabetic Retinopathy

Diabetic Retinopathy is a major cause of vision loss caused by retina lesions, including hard and soft exudates, microaneurysms, and hemorrhages. The development of a computational tool capable of detecting these lesions can assist in the early diagnosis of the most severe forms of the lesions and assist in the screening process and definition of the best treatment form. This paper proposes a computational model based on pre-trained convolutional neural networks capable of detecting fundus lesions to promote medical diagnosis support. The model was trained, adjusted, and evaluated using the DDR Diabetic Retinopathy dataset and implemented based on a YOLOv4 architecture and Darknet framework, reaching an mAP of 11.13% and a mIoU of 13.98%. The experimental results show that the proposed model presented results superior to those obtained in related works found in the literature.

Segmentation in Diabetic Retinopathy using Deeply-Supervised Multiscalar Attention

Diabetic Retinopathy (DR) is a progressive chronic eye disease that leads to irreversible blindness. Detection of DR at an early stage of the disease is crucial and requires proper detection of minute DR pathologies. A novel Deeply-Supervised Multiscale Attention U-Net (Mult-Attn-U-Net) is proposed for segmentation of different DR pathologies viz. Microaneurysms (MA), Hemorrhages (HE), Soft and Hard Exudates (SE and EX). A publicly available dataset (IDRiD) is used to evaluate the performance. Comparative study with four state-of-the-art models establishes its superiority. The best segmentation accuracy obtained by the model for MA, HE, SE are 0.65, 0.70, 0.72, respectively.

CABNet: Category Attention Block for Imbalanced Diabetic Retinopathy Grading

Diabetic Retinopathy (DR) grading is challenging due to the presence of intra-class variations, small lesions and imbalanced data distributions. The key for solving fine-grained DR grading is to find more discriminative features corresponding to subtle visual differences, such as microaneurysms, hemorrhages and soft exudates. However, small lesions are quite difficult to identify using traditional convolutional neural networks (CNNs), and an imbalanced DR data distribution will cause the model to pay too much attention to DR grades with more samples, greatly affecting the final grading performance. In this article, we focus on developing an attention module to address these issues. Specifically, for imbalanced DR data distributions, we propose a novel Category Attention Block (CAB), which explores more discriminative region-wise features for each DR grade and treats each category equally. In order to capture more detailed small lesion information, we also propose the Global Attention Block (GAB), which can exploit detailed and class-agnostic global attention feature maps for fundus images. By aggregating the attention blocks with a backbone network, the CABNet is constructed for DR grading. The attention blocks can be applied to a wide range of backbone networks and trained efficiently in an end-to-end manner. Comprehensive experiments are conducted on three publicly available datasets, showing that CABNet produces significant performance improvements for existing state-of-the-art deep architectures with few additional parameters and achieves the state-of-the-art results for DR grading. Code and models will be available at https://github.com/he2016012996/CABnet.

DRAN: Densely Reversed Attention based Convolutional Network for Diabetic Retinopathy Detection

Diabetic Retinopathy (DR), the complication leading to vision loss, is generally graded according to the amalgamation of various structural factors in fundus photography such as number of microaneurysms, hemorrhages, vascular abnormalities, etc. To this end, Convolution Neural Network (CNN) with impressively representational power has been exhaustively utilized to address this problem. However, while existing multi-stream networks are costly, the conventional CNNs do not consider multiple levels of semantic context, which suffers from the loss of spatial correlations between the aforementioned DR-related signs. Therefore, this paper proposes a Densely Reversed Attention based CNN (DRAN) to leverage the learnable integration of channel-wise attention at multi-level features in a pretrained network for unambiguously involving spatial representations of important DR-oriented factors. Consequently, the proposed approach gains a quadratic weighted kappa of 85.6% on Kaggle DR detection dataset, which is competitive with the state-of-the-arts.

Fusion of Deep Convolutional Neural Networks for Microaneurysm Detection in Color Fundus Images

Microaneurysms (MAs) are common signsof several diseases, appearing as small circular darkish spots in color fundus images. The presence of even a single MA may suggest diseases (e.g. diabetic retinopathy), thus, their reliable recognition is a critical issue in both human clinical practice and computer-aided systems. As for their automatic recognition, deep learning techniques became very popular in the recent years. In this paper, we also apply such deep convolutional neural network (DCNN) based techniques; however, we organize them into a supernetwork with a fusionbased approach. The combination of the member DCNNs is achieved with interconnecting them in a joint fully-connected layer. The advantage of the method is that this large architecture can be trained as a single neural network, and thus, the member DCNNs are also trained with taking the predictions of the other members into consideration. The competitiveness of our approach is also validated with experimental studies, where the ensemble-based system outperformed each member DCNN. As a primary application domain with strong clinical motivation, the methodology was tested for image-level classification. More specifically, a retinal image is divided into subimages to provide the required inputs for the DCNN-based architecture, and the whole image is labeled as a positive case, if the presence of MA is predicted in any of the subimages. Additionally, we also demonstrate how our architecture can be trained to accurately localize MAs with training only the local neighborhoods of the lesions; empirical tests showing solid performance are also enclosed.

Introducing a Novel Layer in Convolutional Neural Network for Automatic Identification of Diabetic Retinopathy

Convolutional neural networks have been widely used for identifying diabetic retinopathy on color fundus images. For such application, we proposed a novel framework for the convolutional neural network architecture by embedding a preprocessing layer followed by the first convolutional layer to increase the performance of the convolutional neural network classifier. Two image enhancement techniques i.e. 1- Contrast Enhancement 2- Contrast-limited adaptive histogram equalization were separately embedded in the proposed layer and the results were compared. For identification of exudates, hemorrhages and microaneurysms, the proposed framework achieved the total accuracy of 87.6%, and 83.9% for the contrast enhancement and contrast-limited adaptive histogram equalization layers, respectively. However, the total accuracy of the convolutional neural network alone without the prreprocessing layer was found to be 81.4%. Consequently, the new convolutional neural network architecture with the proposed preprocessing layer improved the performance of convolutional neural network.

New hierarchical approach for microaneurysms detection with matched filter and machine learning

Microaneurysms are regarded as the first signs of diabetic retinopathy (DR), but the microaneurysms are not clear in the color retinal images, and many researches were studied to detect and locate these lesions. In this paper, a new hierarchical computing-aided diagnosis approach is proposed for the microaneurysms detection by using the multi-scale and multi-orientation sum of matched filter (MMMF) and machine learning, where 37 dimensional features are extracted from each candidate. Furthermore, several classifiers such as the k-nearest neighbor (kNN), local linear discrimination analysis (LLDA) and support vector machine (SVM) are modified to distinguish the true microaneurysms from the false ones, which is a typical unbalanced classification problem. The effectiveness of the proposed method is verified through the training set of a publicly available database, and the experiment results show that the proposed method has better detection performance including the receiver operating characteristic (ROC) curve and the free-response receiver operating characteristic (FROC) curve. Moreover, the proposed method with 37 dimensional features outperforms that with other features and has a sensitivity from 1/8 to 8 with the average of all seven points being 0.286 tested on the same database.