Hybrid Nature Inspired SMO-GBM Classifier for Exudate Classification on Fundus Retinal Images

Background: The diabetic retinopathy can result in loss of vision if not detected in the earlier stages. Exudates are the lesions which play a crucial role in early diagnosis of diabetic retinopathy. The localization of exudates lesions with high values of performance metrics is complicated due to presence of blood vessels and other noisy artifacts. Method: We present computer aided system for classification of retinal fundus images using a novel nature inspired spider monkey optimization for parameter tuning of gradient boosting machines classifier. The image enhancement has been performed with histogram equalization and contourlet transform. The pixels belonging to optic disc region are detected and eliminated using circular Hough transform and Otsu's segmentation method. We have employed Kirsch's matrices for blood vessel detection. The GLCM based feature vector extraction has been employed for textural features. The classification has been performed with hybrid SMO-GBM classifier. Result: We have utilized the STARE database for validation of proposed technique. The proposed system can effectively classify entire image set from test data. The SMO-GBM classifier can further sub-segregate into sub classes with an average accuracy of 97.5%. Conclusion: The proposed approach provides detection and grading of diabetic retinopathy. The abnormality is further categories as soft, moderate and severe. The hybrid SMO-GBM classifier yields a better statistical metrics than the existing exudates classification approaches.     

Automatic non-proliferative diabetic retinopathy screening system based on color fundus image

Background

Non-proliferative diabetic retinopathy is the early stage of diabetic retinopathy. Automatic detection of non-proliferative diabetic retinopathy is significant for clinical diagnosis, early screening and course progression of patients.

Methods

This paper introduces the design and implementation of an automatic system for screening non-proliferative diabetic retinopathy based on color fundus images. Firstly, the fundus structures, including blood vessels, optic disc and macula, are extracted and located, respectively. In particular, a new optic disc localization method using parabolic fitting is proposed based on the physiological structure characteristics of optic disc and blood vessels. Then, early lesions, such as microaneurysms, hemorrhages and hard exudates, are detected based on their respective characteristics. An equivalent optical model simulating human eyes is designed based on the anatomical structure of retina. Main structures and early lesions are reconstructed in the 3D space for better visualization. Finally, the severity of each image is evaluated based on the international criteria of diabetic retinopathy.

Results

The system has been tested on public databases and images from hospitals. Experimental results demonstrate that the proposed system achieves high accuracy for main structures and early lesions detection. The results of severity classification for non-proliferative diabetic retinopathy are also accurate and suitable.

Conclusions

Our system can assist ophthalmologists for clinical diagnosis, automatic screening and course progression of patients.

     

A Multi-Branch Convolutional Neural Network for Screening and Staging of Diabetic Retinopathy Based on Wide-Field Optical Coherence Tomography Angiography

BackgroundDiabetic retinopathy (DR) is one of the major causes of blindness in adults suffering from diabetes. With the development of wide-field optical coherence tomography angiography (WF-OCTA), it is to become a gold standard for diagnosing DR. The demand for automated DR diagnosis system based on OCTA images have been fostered due to large diabetic population and pervasiveness of retinopathy cases.Materials and methodsIn this study, 288 diabetic patients and 97 healthy people were imaged by the swept-source optical coherence tomography (SS-OCT) with 12 mm × 12 mm single scan centered on the fovea. A multi-branch convolutional neural network (CNN) was proposed to classify WF-OCTA images into four grades: no DR, mild non-proliferative diabetic retinopathy (NPDR), moderate to severe NPDR, and proliferative diabetic retinopathy (PDR).ResultsThe proposed model achieved a classification accuracy of 96.11%, sensitivity of 98.08% and specificity of 89.43% in detecting DR. The accuracy of the model for DR staging is 90.56%, which is higher than that of other mainstream convolution neural network models.ConclusionThis technology enables early diagnosis and objective tracking of disease progression, which may be useful for optimal treatment to reduce vision loss.     

A Comprehensive Texture Segmentation Framework for Segmentation of Capillary Non-Perfusion Regions in Fundus Fluorescein Angiograms

Capillary non-perfusion (CNP) in the retina is a characteristic feature used in the management of a wide range of retinal diseases. There is no well-established computation tool for assessing the extent of CNP. We propose a novel texture segmentation framework to address this problem. This framework comprises three major steps: pre-processing, unsupervised total variation texture segmentation, and supervised segmentation. It employs a state-of-the-art multiphase total variation texture segmentation model which is enhanced by new kernel based region terms. The model can be applied to texture and intensity-based multiphase problems. A supervised segmentation step allows the framework to take expert knowledge into account, an AdaBoost classifier with weighted cost coefficient is chosen to tackle imbalanced data classification problems. To demonstrate its effectiveness, we applied this framework to 48 images from malarial retinopathy and 10 images from ischemic diabetic maculopathy. The performance of segmentation is satisfactory when compared to a reference standard of manual delineations: accuracy, sensitivity and specificity are 89.0%, 73.0%, and 90.8% respectively for the malarial retinopathy dataset and 80.8%, 70.6%, and 82.1% respectively for the diabetic maculopathy dataset. In terms of region-wise analysis, this method achieved an accuracy of 76.3% (45 out of 59 regions) for the malarial retinopathy dataset and 73.9% (17 out of 26 regions) for the diabetic maculopathy dataset. This comprehensive segmentation framework can quantify capillary non-perfusion in retinopathy from two distinct etiologies, and has the potential to be adopted for wider applications.     

Mediators of ocular angiogenesis

Angiogenesis is the formation of new blood vessels from pre-existing vasculature. Pathologic angiogenesis in the eye can lead to severe visual impairment. In our review, we discuss the roles of both pro-angiogenic and anti-angiogenic molecular players in corneal angiogenesis, proliferative diabetic retinopathy, exudative macular degeneration and retinopathy of prematurity, highlighting novel targets that have emerged over the past decade.     

Possible new method to improve detection of diabetic retinopathy: Polaroid non-mydriatic retinal photography.

Two retinal cameras (Canon CR2 45NM and CR3 45NM) have recently become available and are capable of producing an instant colour photography of a 45 degree field of retina, including the macula and optic disc, without dilatation of the pupils being necessary. The ability of each camera to detect diabetic retinopathy was compared with that of doctors in diabetic clinics using ophthalmoscopy during busy clinic hours. The CR3 was found to be considerably superior to the CR2 in terms of quality of photograph because it can use a smaller pupil. Overall, the detection rate of the camera was more than four times higher than that of ophthalmoscopy through undilated pupils and more than twice as high as that of ophthalmoscopy through dilated pupils. Lesions missed by ophthalmoscopy but detected by the camera included soft exudates and circinate rings of hard exudates, sometimes encroaching on the macula. Though various aspects of this system of screening for diabetic retinopathy, in particular its ability to detect new retinal vessels, have not yet been assessed, the system may prove beneficial in the detection and monitoring of diabetic retinopathy.     

Assessment of non-mydriatic fundus photography in detection of diabetic retinopathy.

Non-mydriatic retinal photography with later interpretation of the photographs was assessed as a screening method for the detection of diabetic retinopathy; when compared with an ophthalmologist''s clinical assessment in a random group of 62 diabetic patients it was accurate (false negative 6.8%, false positive 2%) and sensitive (sensitivity 96%, specificity 98%). The assessment of further management required based on analysis of the photographs was 96.5% in agreement with the further management suggested by the ophthalmologist after direct clinical assessment of the patient. If this technique were used to screen patients in a typical diabetic clinic the predicted positive accuracy rate would be 84% and the predicted negative accuracy rate 99.5%.     

Diabetic retinopathy in the Eastern Morocco: Different stage frequencies and associated risk factors

Diabetes is a major cause of morbidity and mortality worldwide. It can affect many organs and, over time, leads to serious complications. Diabetic retinopathy (DR), a specific ocular complication of diabetes, remains the leading cause of vision loss and vision impairment in adults. This work is the first in Eastern Morocco aimed at identifying the different stages of DR and to determine their frequencies and associated risk factors. It is a case-control study conducted from December 2018 to July 2019 at the ophthalmology department of Al-Irfane Clinic (Oujda). Data were obtained from a specific questionnaire involving 244 diabetic patients (122 cases with retinopathy vs 122 controls without retinopathy). All results were analyzed by the EPI-Info software. This study shows a predominance of proliferative diabetic retinopathy (PDR) with 57.4% of cases (uncomplicated proliferative diabetic retinopathy (UPDR): 23.8%; complicated proliferative diabetic retinopathy (CPDR): 33.6%). The non-proliferative diabetic retinopathy (NPDR) represents 42.6% (minimal NPDR: 8.2%; moderate NPDR: 26.2%; severe NPDR: 8.2%). The determinants of DR were insulin therapy, high blood pressure, poor glycemic control and duration of diabetes. Regarding the chronological evolution, retinopathy precedes nephropathy. Diabetic nephropathy (DN) was present in 10.6% of cases especially in patients with PDR. In summary, the frequency of PDR was higher than that of NPDR. DR appears before DN with a high frequency of DN in patients with PDR. Good glycemic control and blood pressure control, as well as early diagnosis are the major preventive measures against DR.     

Prevalence of retinopathy in a diabetic clinic.

In a prospective study lasting 14 months an attempt was made to measure the visual acuity and examine the fundi, after mydriasis, of all patients attending the diabetic clinic of a district general hospital. Of 704 patients, 160 (22.7%) had some evidence of retinopathy, and 52 (7.4%) of these were already attending an ophthalmologist. A further 18 (2.6%) were known to have retinopathy and were being followed up in the diabetic clinic. Ninety (12.8%) new patients with diabetic retinopathy were discovered. Most had minimal changes, but 30 (4.3%) were considered to have changes severe enough to be referred to an ophthalmologist. Twenty-two (2.1%) underwent, or were awaiting, photocoagulation, and half of these had had no visual symptoms when first seen. Although some of these patients were already being treated or observed for retinopathy, it is encouraging that relatively few new patients needing treatment for retinopathy were discovered. If retinopathy could be detected early enough physicians might be able to deal with it and so ease pressure on ophthalmological services.     

Effects of intraocular or systemic administration of neutralizing antibody against vascular endothelial growth factor on the murine experimental model of retinopathy

Vascular endothelial growth factor (VEGF), the strongest known angiogenic cytokine and also a potent enhancer of vascular permeability, is closely associated with diabetic ocular complications and other intraocular neovascular diseases. The therapeutic effect of VEGF-neutralizing antibody on oxygen-induced retinopathy in an experimental murine model of proliferative retinopathy was investigated. Intraocular and systemic injection of the antibody resulted in 46% and 18% reductions in the number of nuclei of newly formed vessels of this model, respectively. The results demonstrated that a neutralizing antibody against VEGF was highly effective in the treatment of intraocular neovascularization and suggested possible modes of therapy in human intraocular neovascular diseases, including diabetic proliferative retinopathy.     

Automated quantification of superficial retinal capillaries and large vessels for diabetic retinopathy on optical coherence tomographic angiography

Optical coherence tomography angiography (OCTA) is a relatively new technique with capillary‐level resolution, which has shown great potential for the diagnosis of diabetic retinopathy (DR). A fully automatic algorithm for the quantitative measurement of microcirculatory changes in sight‐threatening DR is presented. The foveal avascular zone (FAZ) segmentation was improved with a graph‐theoretic method and the large vessels and capillaries were separately identified and analyzed. The method was evaluated in healthy and diabetic eyes with various stages of retinopathy. Results showed that, compared with the healthy group, the diabetic group showed a significantly larger large vessel density, but a significantly smaller capillary density (P < .001). Circularity of FAZ was significantly smaller while nonperfusion area was significantly larger in the diabetic group. The combined variable of all image metrics reached an area under the ROC of 0.853 (95% CI, 0.784‐0.923) for mild to moderate nonproliferative DR and 0.950 (95% CI, 0.922‐0.979) for proliferative DR. Microvascular and FAZ changes with various DR stages can be accurately delineated using the developed automatic program. Quantitative metrics on OCTA serve as potential biomarkers for the staging of DR.     

Adeno-Associated Virus Mediated Delivery of a Non-Membrane Targeted Human Soluble CD59 Attenuates Some Aspects of Diabetic Retinopathy in Mice

Diabetic retinopathy is the leading cause of visual dysfunction in working adults and is attributed to retinal vascular and neural cell damage. Recent studies have described elevated levels of membrane attack complex (MAC) and reduced levels of membrane associated complement regulators including CD55 and CD59 in the retina of diabetic retinopathy patients as well as in animal models of this disease. We have previously described the development of a soluble membrane-independent form of CD59 (sCD59) that when delivered via a gene therapy approach using an adeno-associated virus vector (AAV2/8-sCD59) to the eyes of mice, can block MAC deposition and choroidal neovascularization. Here, we examine AAV2/8-sCD59 mediated attenuation of MAC deposition and ensuing complement mediated damage to the retina of mice following streptozotocin (STZ) induced diabetes. We observed a 60% reduction in leakage of retinal blood vessels in diabetic eyes pre-injected with AAV2/8-sCD59 relative to negative control virus injected diabetic eyes. AAV2/8-sCD59 injected eyes also exhibited protection from non-perfusion of retinal blood vessels. In addition, a 200% reduction in retinal ganglion cell apoptosis and a 40% reduction in MAC deposition were documented in diabetic eyes pre-injected with AAV2/8-sCD59 relative to diabetic eyes pre-injected with the control virus. This is the first study characterizing a viral gene therapy intervention that targets MAC in a model of diabetic retinopathy. Use of AAV2/8-sCD59 warrants further exploration as a potential therapy for advanced stages of diabetic retinopathy.     

Potent antiemetic effects of the new serotonin antagonists.

Diabetic retinopathy progresses through three distinct stages. A rational approach to management is based on an understanding of the pathophysiology of each stage. Based on the results of national multicentered clinical trials of laser photocoagulation and other treatments, advances in our understanding of the pathogenesis and treatment can now make a dramatic impact on blindness in the diabetic population: Panretinal laser photocoagulation treatment can reduce the risk of vision loss from high-risk proliferative diabetic retinopathy by at least 50%. Laser photocoagulation treatment of clinically significant diabetic macular edema can reduce the risk of vision loss by more than 50%. Vitrectomy can restore useful vision to some patients with severe diabetic retinopathy and vitreous hemorrhage with or without an accompanying traction retinal detachment. Diabetes 2000 is a new project sponsored by the American Academy of Ophthalmology, the goal of which is to eliminate preventable blindness from diabetes by the year 2000. As its name implies, Diabetes 2000 will be a long-term project aimed at a specific disease--diabetic retinopathy and its complications. It will provide the latest research findings to ophthalmologists and primary care physicians as the first priority, followed by the education of patients and the general public. Recent advances and treatment guidelines for the medical and surgical treatment of diabetic eye disease will be emphasized through the continuing education of ophthalmologists, other physicians, and allied health professionals. In later phases, educational programs for diabetic persons and the public will be developed.(ABSTRACT TRUNCATED AT 250 WORDS)     

Suppression of protein kinase C-ζ attenuates vascular leakage via prevention of tight junction protein decrease in diabetic retinopathy

To investigate the effect of protein kinase C (PKC)-ζ inhibition on vascular leakage in diabetic retinopathy, streptozotocin-induced diabetic mice were intravitreously injected with siPKC-ζ. According to the fluorescein angiography of the retinal vessels, suppression of PKC-ζ effectively attenuated vascular leakage in diabetic retina. Further evaluation on the retina with western blot analysis and immunohistochemistry revealed accompanying restoration of tight junction proteins on retinal vessels. As two major contributors to vascular leakage in diabetic retinopathy, vascular endothelial growth factor (VEGF) and advanced glycation end products (AGEs) were investigated on the tight junction protein expression in endothelial cells. Inhibition of PKC-ζ attenuated VEGF-induced decrease of tight junction proteins and accompanying hyperpermeability in human retinal microvascular endothelial cells (HRMECs). PKC-ζ inhibition also attenuated AGE-induced decrease of tight junction proteins in HRMECs. Our findings suggest that inhibition of PKC-ζ could be an alternative treatment option for compromised blood-retinal barrier in diabetic retinopathy.     

Automatic wavelet-based retinal blood vessels segmentation and vessel diameter estimation

Automatic detection of retinal blood vessels and measurement of vessel diameter are important steps in the computer aided diagnosis in ophthalmology. Here, we present a new multi-scale vessel enhancement method based on complex continuous wavelet transform (CCWT). The parameters of CCWT are optimized to represent line structures in all directions and separate them from simple edges. The final vessel network is obtained by applying an adaptive histogram-based thresholding process along with a proper length filtering method. An efficient circular structure operator is employed on the centerline of vessels to estimate their diameters. The performance of the proposed method is measured on the publicly available DRIVE and STARE databases and compared with several state-of-the-art methods as well as second observer. The proposed method shows much higher accuracy (95%) and sensitivity (79%) in the same range of specificity (97%). The predictive value of it is higher than 72.9%. The vessel diameter estimation process also shows lower root mean square error compared to the existing methods and second observer.     

Pituitary Ablation for Diabetic Retinopathy

Pituitary ablation was performed on 26 patients with advanced diabetic retinopathy by pituitary stalk section (nine patients) or transphenoidal hypophysectomy (17 patients). After a latent period varying from six to 12 months, retinal hemorrhages disappeared, vitreous hemorrhages stopped and new vessels regressed. Eight of 10 patients followed up for more than a year had a complete remission, although in one of two patients whose ablation was incomplete the retinopathy remained active. Preoperative visual acuity levels were preserved in 47 of 50 eyes of patients followed up for a mean period of 12 months. Three patients have died, two from causes related to their operation. Twenty of the survivors have continued in their usual occupation. Pituitary ablation is effective in preventing the progressive visual deterioration usually associated with advanced diabetic retinopathy.     

A Higher Serum Calcium Level is an Independent Risk Factor for Vision-Threatening Diabetic Retinopathy in Patients with Type 2 Diabetes: Cross-Sectional and Longitudinal Analyses

ObjectiveAn elevated serum calcium level is associated with a higher risk of type 2 diabetes (T2D), but its role in microvascular complications remains unclear. This study was conducted to investigate the association between serum calcium levels and vision-threatening diabetic retinopathy (VTDR).MethodsThis study employed a cross-sectional and longitudinal design. The cross-sectional part included all patients treated for T2D at Shanghai General Hospital between 2007 and 2016, while the longitudinal part involved an overlapping cohort of diabetic patients without VTDR who were followed from their admission until December 2019. Multivariable logistic and Cox proportional hazard regression analyses were performed, respectively. VTDR was defined as severe nonproliferative diabetic retinopathy, proliferative diabetic retinopathy, or clinically significant macular edema.ResultsA total of 3269 patients were included in the cross-sectional analysis, and 649 patients were included in the longitudinal analysis. In the cross-sectional analysis, higher corrected serum calcium (odds ratio: 1.31 per 0.1 mmol/L, 95% confidence interval: 1.16-1.49), younger age, longer diabetes duration, albuminuria, impaired renal function, and lower serum magnesium were independently associated with VTDR. In the longitudinal analysis, 95 subjects developed VTDR during follow-up (9.7 years, interquartile range: 7.4-10.9 years). Higher corrected serum calcium (hazard ratio: 1.38 per 0.1 mmol/L, 95% confidence interval: 1.10-1.72), younger age, longer diabetes duration, sub-VTDR, albuminuria, lower serum magnesium, and higher glycated hemoglobin were identified as independent risk factors for VTDR.ConclusionsA higher serum calcium level may be an independent risk factor for VTDR in patients with T2D.     

Proliferative diabetic retinopathy: treatment with xenon-arc photocoagulation. Interim report of multicentre randomised controlled trial.

One hundred patients with symmetrical proliferative diabetic retinopathy had one eye randomly chosen for treatment with xenonarc photocoagulation while the other was left untreated as a control. Patients were subdivided into those with new vessels on both optic discs and those with only peripheral new vessels. In patients with new vessels on the optic discs the vision of the untreated eyes deteriorated more than that of the treated eyes and the difference in deterioration was significant after one, two, and three years. There was no such difference in patients who had only peripheral new vessels. Eighteen patients had become blind in one or both eyes by the last assessment, but only one patient became blind in the treated eye without concomitant blindness in the untreated eye. Thirteen were blind only in the untreated eye. Both photographic and ophthalmoscopic examinations showed that new vessels on the disc regressed more in the treated eyes than in the untreated ones. As some forms of diabetic retinopathy are now treatable, early diagnosis and evaluation is increasingly important.     

RhoB antibody alters retinal vascularization in models of murine retinopathy

Neovascularization in cancer or retinopathy is driven by pathological changes that foster abnormal sprouting of endothelial cells. Mouse genetic studies indicate that the stress-induced small GTPase RhoB is dispensable for normal physiology but required for pathogenic angiogenesis. In diabetic retinopathy, retinopathy of prematurity (ROP) or age-related wet macular degeneration (AMD), progressive pathologic anatomic changes and ischemia foster neovascularization are characterized by abnormal sprouting of endothelial cells. This process is driven by the angiogenic growth factor VEGF, which induces and supports the formation of new blood vessels. While injectable biologics targeting VEGF have been used to treat these pathological conditions, many patients respond poorly, prompting interest in other types of mechanism-based therapy. Here we report the preclinical efficacy of a monoclonal antibody that specifically targets RhoB, a signaling molecule that is genetically dispensable for normal physiology but required for pathogenic retinal angiogenesis. In murine models of proliferative retinal angiogenesis or oxygen-induced retinopathy, administering a monoclonal RhoB antibody (7F7) was sufficient to block neoangiogenesis or avascular pathology, respectively. Our findings offer preclinical proof of concept for antibody targeting of RhoB to limit diabetic retinopathy, ROP or wet AMD and perhaps other diseases of neovasculogenesis such as hemangioma or hemangiosarcoma nonresponsive to existing therapies.     

A Generic Approach for Efficient Detection of Vascular Structures

Vascular segmentation is often required in medical image analysis for various imaging modalities. Despite the rich literature in the field, the proposed methods need most of the time adaptation to the particular investigation and may sometimes lack the desired accuracy in terms of true positive and false positive detection rate. This paper proposes a general method for vascular segmentation based on locally connected filtering applied in a multiresolution scheme. The filtering scheme performs progressive detection and removal of the vessels from the image relief at each resolution level, by combining directional 2D-3D locally connected filters (LCF). An important property of the LCF is that it preserves (positive contrasted) structures in the image if they are topologically connected with other similar structures in their local environment. Vessels, which appear as curvilinear structures, can be filtered out by an appropriate LCF set-up which will minimally affect sheet-like structures. The implementation in a multiresolution framework allows dealing with different vessel sizes. The outcome of the proposed approach is illustrated on several image modalities including lung, liver and coronary arteries. It is shown that besides preserving high accuracy in detecting small vessels, the proposed technique is less sensitive with respect to noise and the presence of pathologies of positive-contrast appearance on the images. The detection accuracy is compared with a previously developed approach on the 20 patient database from the VESSEL12 challenge.