Optic Disc Cupping Research Articles

Disc damage likelihood scale recognition for Glaucoma detection

Glaucoma is a visual disorder, which is one of the significant driving reason for visual impairment. Glaucoma leads to frustrate the visual information transmission to the brain. Dissimilar to other eye illness such as myopia and cataracts. The impact of glaucoma can’t be cured; The Disc Damage Likelihood Scale (DDLS) can be used to assess the Glaucoma. The proposed methodology suggested simple method to extract Neuroretinal rim (NRM) region then dividing the region into four sectors after that calculate the width for each sector and select the minimum value to use it in DDLS factor. The feature was fed to the SVM classification algorithm, the DDLS successfully classified Glaucoma disease with 70% percentage; moreover, when the dimensions of both Optic Disc(OD) and Optic Cup(OC) were used as additional features the accuracy rate raised to 91%.

Axial Length Changes Following Surgical Intervention in Children With Primary Congenital Glaucoma.

Primary congenital glaucoma (PCG) is a challenging condition to diagnose, treat and effectively monitor. Serial assessment of intraocular pressure (IOP), optic disc cupping, refraction, and axial length (AxL) after surgery are useful to assess disease control. This study aimed to evaluate AxL changes in relation to IOP changes following glaucoma surgery in children with PCG. We retrospectively studied AxL changes in children with PCG undergoing surgery. Eyes of children aged ≤ 4 years that did not have prior ocular surgery and that underwent at least one glaucoma surgery during the course of follow-up between June 2014 and July 2018, were included. The effect of change in IOP on change in AxL was estimated using linear mixed effects models. A total of 105 eyes (of 72 children) with PCG underwent glaucoma surgery representing 26.4% (105/397) eyes. The mean ± SD age of children at baseline was 3.53± 4.04 months. At baseline, the mean IOP and AxL were 26.63 ± 9.57 mmHg and 21.67 ± 1.82mm, respectively. During the course of follow-up post-surgery, the IOP decreased by a mean of 7.25 ± 12.08 mmHg while the AxL increased by a mean of 0.70 ± 1.40mm. A multivariable mixed effects linear regression revealed that change in AxL was significantly associated with change in IOP (p=0.030) and time since first surgery (p<0.001). A substantial reduction in IOP (≥35 mmHg) was needed at 3 months post-surgery, for AxL to regress. In children with PCG who undergo glaucoma surgery, change in IOP significantly influences change in AxL. For AxL to regress, a substantial reduction in IOP is needed post-surgery.

A Survey of the Relationship Between Serum Cholesterol and Triglyceride to Glaucoma: A Case Control Study

Purpose/Aim: primary Open-angle glaucoma (POAG) is an asymptomatic, progressive optic neuropathy characterized by enlarging optic disc cupping and visual field loss. POAG is a major cause of blindness and is characterized by progressive degeneration of the optic nerve and is usually associated with elevated intraocular pressure. Regarding the fact that dyslipidemia has a relationship with some ophthalmic diseases such as cataract,it sounds that the same relationship also exists with POAG. Therefore, it was decided to study the relationship between serum cholesterol and triglyceride to POAG. Methods: The present study was done on 40 primary Open-angle glaucoma (POAG) patients (cases) and 40 healthy individuals (controls). In order to diagnose POAG, Intra ocular pressure (IOP) was measured by means of Applanation tonometry, and then was confirmed through perimetry and ophthalmoscopes. The controls were patients’ attendants who had referred to Birjand Valli-e-asr clinics but did not have POAG. After a fasting of 12 hours, blood samples were derived to determine serum level of triglyceride and cholesterol levels. Then, the results of the tests together with the participants’ demographic information were individually registered in a questionnaire. The obtained data was analyzed by means of SPSS software (version 15) and statistical tests including T-test and Chi- square. Results: Mean levels of cholesterol (211.18± 51.91mg/dl in cases, 162.38±39.56 mg/dl in controls) and triglyceride (165.92±88.58 mg/dl in cases, 99.46± 43.08 mg/dl in control) were significantly higher in cases than in controls. Hypercholesterolemia and hypertriglyceridemia were significantly higher in the cases compared with the controls. There was a positive association between POAG and dyslipidemia (OR=7.14 [95% CI: 2.3-22.2] for Hypercholesterolemia and OR=16.9 [95% CI: 2.1-14.8] for hypertriglyceridemia. Conclusion: Hyperlipidemia can be a risk factor of getting POAG.

Segmentation technique and dynamic ensemble selection to enhance glaucoma severity detection

The severity of glaucoma can be observed by categorising glaucoma diseases into several classes based on a classification process. The two most suitable parameters, cup-to-disc ratio (CDR) and peripapillary atrophy (PPA), which are commonly used to identify glaucoma are utilized in this study to strengthen the classification. First, an active contour snake (ACS) is employed to retrieve both optic disc (OD) and optic cup (OC) values, which are required to calculate the CDR. Moreover, Otsu segmentation and thresholding techniques are used to identify PPA, and the features are then extracted using a grey-level co-occurrence matrix (GLCM). An advanced segmentation technique, combined with an improved classifier called dynamic ensemble selection (DES), is proposed to classify glaucoma. Because DES is generally used to handle an imbalanced dataset, the proposed model is expected to detect glaucoma severity and determine the subsequent treatment accurately. The proposed model obtains a higher mean accuracy (0.96) than the deep learning-based U-Net (0.90) when evaluated using three datasets of 250 retinal fundus images (200 training, 50 testings) based on the 5-fold cross-validation scheme.

Psychosocial indicators of primary congenital glaucoma and filtering surgeries impacts on children and family's quality of life.

To identify psychosocial indicators and evaluate the filtering surgeries impacts on quality of life (QOL) of children with primary congenital glaucoma (PCG) and their family. Parents of children with bilateral PCG who underwent filtering surgery were included. Data were collected through: (a) psychological inquiry to determine psychosocial indicators; (b) Children's Visual Function Questionnaire (CVFQ). The associations between the CVFQ scores and visual acuity, intraocular pressure, optic disk cupping, horizontal corneal diameter, axial length, number of surgeries, and hypotensive eye drops were investigated. The statistical significance level was considered as p ⩽ 0.05. The mean age of the nine mothers interviewed and their children (six boys and three girls) was 29 ± 5 years and 35 ± 18 months, respectively. The psychosocial indicators determined were: emotional impact of the diagnosis, disease knowledge, mother and family's feelings on facing the surgical treatment, surgical result comprehension, treatment adherence, child's emotional and behavioral reactions, social support, and future expectations. In CVFQ analysis, QOL score was strongly associated with visual acuity (r = -0.79; p = 0.01). Besides the treatment score was correlated to intraocular pressure (r = -0.68; p < 0.05), optic disk cupping (r = -0.85; p = 0.03), and corneal diameter (r = -0.69; p = 0.02). Correlations were not found for number of surgeries and eye drops. This study confirmed the PCG surgical treatment impact on QOL and determined psychosocial indicators which may favor the suitable actions in psychological treatment and follow-up of the children and families.

Joint optic disc and cup segmentation based on multi-scale feature analysis and attention pyramid architecture for glaucoma screening

Automatic segmentation of optic disc (OD) and optic cup (OC) is an essential task for analysing colour fundus images. In clinical practice, accurate OD and OC segmentation assist ophthalmologists in diagnosing glaucoma. In this paper, we propose a unified convolutional neural network, named ResFPN-Net, which learns the boundary feature and the inner relation between OD and OC for automatic segmentation. The proposed ResFPN-Net is mainly composed of multi-scale feature extractor, multi-scale segmentation transition and attention pyramid architecture. The multi-scale feature extractor achieved the feature encoding of fundus images and captured the boundary representations. The multi-scale segmentation transition is employed to retain the features of different scales. Moreover, an attention pyramid architecture is proposed to learn rich representations and the mutual connection in the OD and OC. To verify the effectiveness of the proposed method, we conducted extensive experiments on two public datasets. On the Drishti-GS database, we achieved a Dice coefficient of 97.59%, 89.87%, the accuracy of 99.21%, 98.77%, and the Averaged Hausdorff distance of 0.099, 0.882 on the OD and OC segmentation, respectively. We achieved a Dice coefficient of 96.41%, 83.91%, the accuracy of 99.30%, 99.24%, and the Averaged Hausdorff distance of 0.166, 1.210 on the RIM-ONE database for OD and OC segmentation, respectively. Comprehensive results show that the proposed method outperforms other competitive OD and OC segmentation methods and appears more adaptable in cross-dataset scenarios. The introduced multi-scale loss function achieved significantly lower training loss and higher accuracy compared with other loss functions. Furthermore, the proposed method is further validated in OC to OD ratio calculation task and achieved the best MAE of 0.0499 and 0.0630 on the Drishti-GS and RIM-ONE datasets, respectively. Finally, we evaluated the effectiveness of the glaucoma screening on Drishti-GS and RIM-ONE datasets, achieving the AUC of 0.8947 and 0.7964. These results proved that the proposed ResFPN-Net is effective in analysing fundus images for glaucoma screening and can be applied in other relative biomedical image segmentation applications.

GDCSeg-Net: general optic disc and cup segmentation network for multi-device fundus images.

Accurate segmentation of optic disc (OD) and optic cup (OC) in fundus images is crucial for the analysis of many retinal diseases, such as the screening and diagnosis of glaucoma and atrophy segmentation. Due to domain shift between different datasets caused by different acquisition devices and modes and inadequate training caused by small sample dataset, the existing deep-learning-based OD and OC segmentation networks have poor generalization ability for different fundus image datasets. In this paper, adopting the mixed training strategy based on different datasets for the first time, we propose an encoder-decoder based general OD and OC segmentation network (named as GDCSeg-Net) with the newly designed multi-scale weight-shared attention (MSA) module and densely connected depthwise separable convolution (DSC) module, to effectively overcome these two problems. Experimental results show that our proposed GDCSeg-Net is competitive with other state-of-the-art methods on five different public fundus image datasets, including REFUGE, MESSIDOR, RIM-ONE-R3, Drishti-GS and IDRiD.

Optic Disc and Optic Cup Segmentation for Glaucoma Detection from Blur Retinal Images Using Improved Mask-RCNN

Glaucoma is a fatal eye disease that harms the optic disc (OD) and optic cup (OC) and results into blindness in progressed phases. Because of slow progress, the disease exhibits a small number of symptoms in the initial stages, therefore causing the disease identification to be a complicated task. So, a fully automatic framework is mandatory, which can support the screening process and increase the chances of disease detection in the early stages. In this paper, we deal with the localization and segmentation of the OD and OC for glaucoma detection from blur retinal images. We have presented a novel method that is Densenet-77-based Mask-RCNN to overcome the challenges of the glaucoma detection. Initially, we have performed the data augmentation step together with adding blurriness in samples to increase the diversity of data. Then, we have generated the annotations from ground-truth (GT) images. After that, the Densenet-77 framework is employed at the feature extraction level of Mask-RCNN to compute the deep key points. Finally, the calculated features are used to localize and segment the OD and OC by the custom Mask-RCNN model. For performance evaluation, we have used the ORIGA dataset that is publicly available. Furthermore, we have performed cross-dataset validation on the HRF database to show the robustness of the presented framework. The presented framework has achieved an average precision, recall, F-measure, and IOU as 0.965, 0.963, 0.97, and 0.972, respectively. The proposed method achieved remarkable performance in terms of both efficiency and effectiveness as compared to the latest techniques under the presence of blurring, noise, and light variations.

THE VALUE OF VISUAL EVOKED POTENTIAL TESTING IN OPERATED CASES OF PRIMARY CONGENITAL GLAUCOMA

PCG is a hereditary childhood glaucoma that is not associated with other ocular, systemic congenital anomalies or acquired conditions it accounts for 26–29% of childhood blindness in Egypt.Glaucoma defined according to CGRN classification as the presence of two of these factors: elevated IOP> 21, optic disc cupping and corneal findings.PCG commonly presents bilaterally. The classical triad of symptoms is epiphora, photophobia and blepharospasm and the most common presenting sign is Cloudy corneaTreatment include medical treatment which role is known to be only supportive to decrease the corneal opacity before surgery or as an adjunct to achieve maximum IOP reduction postoperative Surgical treatment in the mild form of congenital glaucoma is generally successful, no matter which procedure is chosen but in the moderate and severe forms combined trabeculotomy-trabeculectomy with adjunctive mitomycin C results in better IOP controlFlash VEP is used in uncooperative patients and its latencies are more variable than the pattern reversal type. Therefore, it only tests continuity of the visual pathways from the optic nerve to the occipital cortex.

Change of ocular parameters in children with large cup-to-disc ratio and interocular cup-to-disc ratio asymmetry.

Enlarged optic disc cupping and interocular cup-to-disc ratio (CDR) asymmetry are often important indicators of glaucoma. Clinically, we occasionally encounter children with large CDR and interocular CDR asymmetry during vision screening. This study aimed to report longitudinal change of ocular parameters in children with large cup-to-disc ratio (CDR) and interocular CDR asymmetry. This was a retrospective, observational case series of 160 eyes of 160 children with large CDR who visited a tertiary eye center from January 2010 to June 2016. Average CDR ≥ 0.6 were considered large CDR values, and CDR asymmetry was defined as an interocular difference ratio value greater than 0.2. All included patients showed interocular pressure (IOP) < 21mmHg at least three ophthalmic examinations conducted at total intervals of at least 30months. The mean age of children included in the study was 7.14 ± 2.42years, with a follow-up period of 54.46 ± 19.82months. Changes in refractive error and axial length were significantly different between initial and final examination (p < 0.001). However, optic nerve head (ONH) analysis and retinal nerve fiber layer (RNFL) and macular ganglion cell-inner plexiform layer (mGCIPL) thicknesses were not significantly different between initial and final examination. In interocular comparisons of patients with CDR asymmetry, changes of refractive error, axial length, ONH analysis, and RNFL and mGCIPL thickness were not significantly different between the two eyes. There were no significant differences in the changes of ONH analysis, and RNFL and mGCIPL thicknesses in children with large CDR, or those with interocular CDR asymmetry over the study period. Our results provide helpful information for the establishment of guidelines for managing children with large CDR and interocular CDR asymmetry.

Visual outcomes and associated factors of primary congenital glaucoma in children.

We evaluated the long-term visual outcomes in children with primary congenital glaucoma and determined the factors associated with the final visual outcomes. Medical records of children with primary congenital glaucoma between 2005 and 2016, seen at Seoul National University Children's Hospital in South Korea, were reviewed. The minimum follow-up period after surgery for primary congenital glaucoma was 3years. Visual acuity (VA) was categorized into good (≧20/70) and poor (< 20/70). Factors including age, VA, refractive errors, intraocular pressure (IOP), laterality, and cup-to-disc (C/D) ratio were compared between the groups. A total of 71 eyes of 44 patients were included. The patients' age at the time of surgery was 14.7 ± 12.2months. The mean IOP was 28.3 ± 7.0mmHg. During 6.7 ± 2.7years of mean follow-up after surgery, 39 eyes (54.9%) needed occlusion treatment. After occlusion, patients with lower IOP values, lesser additional surgeries, reversal of optic disc cupping, and better initially measured VA achieved a better visual outcome. At the final assessment, the mean age was 7.8 ± 2.6years, and the mean VA gain was 15.0 ± 19.4 letters. There were 44 eyes (62.0%) with VA ≧20/70. In children with primary congenital glaucoma, IOP control and the optic disc configuration over time are important factors associated with visual outcome. Regular follow-up and correction of refractive errors-along with occlusion for those with difference in VA between the two eyes-might be helpful for achieving better visual outcomes.

An automatic recognition of glaucoma in fundus images using deep learning and random forest classifier

Glaucoma is perpetual damage of optic nerves which causes fractional or complete visual misfortune. The fundamental reason for this illness is the increment of the intra-ocular pressure inside the eye which harms the optic nerve. Retinal images give indispensable data about an eye’s wellbeing. Based on progressions in retinal images technology, it is conceivable to create frameworks that can investigate these retinal imageries for better determination. This work presents a glaucoma recognition technique by estimating CDR (Cup to Disc Ratio) from fundus images. The size of the optic disc and optic cup is utilized to distinguish the presence of glaucoma. Therefore, the segmentation of the optic disc and optic cup is the primary step in glaucoma recognition. Decreasing the number of features and reducing the error are the two clashing destinations. The proposed glaucoma recognition technique comprises image acquisition, feature extraction, and glaucoma evaluation stages. The contrast enhancement operation is performed in image acquisition. While boundaries of OD (Optic Disc) and OC (optic cup) are segmented in the feature extraction stage and it is performed by utilizing Au-Net. Then CDR ratio of an abused image is computed to assess glaucoma in the images. Thereafter, a random forest classifier has been used to classify the glaucomatous images based on the CDR values. The performance of the proposed method has been evaluated with different techniques such as Deformable U-Net, Full-Deformable U-Net, and Original U-Net. From the outcomes, it can be noticed that the proposed method gives better performance in terms of classification accuracy of 99% and 14% of segmentation accuracy has been compared when compared with Original U-Net.

Joint optic disc and optic cup segmentation based on boundary prior and adversarial learning.

The most direct means of glaucoma screening is to use cup-to-disc ratio via colour fundus photography, the first step of which is the precise segmentation of the optic cup (OC) and optic disc (OD). In recent years, convolution neural networks (CNN) have shown outstanding performance in medical segmentation tasks. However, most CNN-based methods ignore the effect of boundary ambiguity on performance, which leads to low generalization. This paper is dedicated to solving this issue. In this paper, we propose a novel segmentation architecture, called BGA-Net, which introduces an auxiliary boundary branch and adversarial learning to jointly segment OD and OC in a multi-label manner. To generate more accurate results, the generative adversarial network is exploited to encourage boundary and mask predictions to be similar to the ground truth ones. Experimental results show that our BGA-Net system achieves state-of-the-art OC and OD segmentation performance on three publicly available datasets, i.e., the Dice scores for the optic disc/cup on the Drishti-GS, RIM-ONE-r3 and REFUGE datasets are 0.975/0.898, 0.967/0.872 and 0.951/0.866, respectively. In this work, we not only achieve superior OD and OC segmentation results, but also confirm that the values calculated through the geometric relationship between the former two are highly related to glaucoma.

An Analysis Of Segmentation And Glaucoma Diagnosis By Deep Learning Technique Via Retinal Fundus Images

A leading retinal disease was Glaucoma. It damages the eye due to the intraocular pressure on the eye. When glaucoma was untreated left, it would led to the loss of vison by affecting the ONH. (optic nerve head) The glaucoma progression was investigated on the retina of eye through ophthalmologist. This method was very boring and it enhance more time to do by human beings. Therefore, this tiuuse was the right problem which may be solved by diagnosing automatically with the support of the approach of deep learning process. CNNs are suitable to identify the solution for this problem. It may extract different stages of information from the input picture and motivates to variate among glaucomica and non-glaucomic pictures. This examined paper presents an effective glaucoma structure to fragment the optic disc and cup to identify the CDR value. The glaucoma was achieved by deep learing process along with the novel Convolutional Neural Network. An examined method uses 2 similar Convolutional neural network architecture to extract the optic disc and otpic cup to get perfect output. This deisgn was tested and trained on DRISHRI –GS data base, that was available publicly and an exact value of 97% for OC and 98% for OD segmentation was investigated.

Incidence of non-glaucomatous ocular disease in patients with asymmetric optic disc cupping.

To evaluate the incidence of non-glaucomatous ocular disease in patients with asymmetric optic disc cupping. A retrospective case series, including consecutive patients with cup-to-disc ratio (CDR) asymmetry greater than 0.2. All patients underwent a complete neuro-ophthalmological examination, automated perimetry with the Humphrey 24-2 visual fields program. Retinal nerve fibre layer thickness was measured by optical coherence tomography (OCT). The results of neuroimaging, macular OCT and blood tests were recorded as well. Patients were assigned a diagnosis of glaucomatous optic neuropathy (GON) or non-glaucomatous disease (NGD). The main outcome measure was the rate of non-glaucomatous ocular disease. A total of 120 (67 males) patients with a mean age of 71.1 ± 12.5years met the inclusion criteria and were included in this study. The mean asymmetry in CDR between the eyes was 0.3 ± 0.13 (range, 0.2-0.9). Twenty patients (16.6%) had a visual field defect not typical for glaucoma and positive relative afferent pupillary defect was found in 24 patients (20%). Six patients were found to have newly diagnosed non-glaucomatous ocular disease: maculopathy in three patients, retinopathy in one patient and traumatic optic neuropathy in two patients. Patients with NGD were significantly younger than the patients with GON (59.8 ± 23.3 vs. 71.3 ± 11.5years, P = 0.001). Optic disc pallor was found in 4/93 patients with glaucoma compared to 3/6 with newly diagnosed non-glaucomatous disease (4.7% vs. 50.0%, P = 0.03). Asymmetric optic disc cupping can be associated with non-glaucomatous disease and may warrant neuro-ophthalmological evaluation, especially in younger patients or those with optic disc pallor.

Automated segmentation of optic disc and optic cup for glaucoma assessment using improved UNET++ architecture

Glaucoma is one of the leading cause of blindness for over 60 million people around the world. Since a cure for glaucoma doesn’t yet exist, early screening and diagnosis become critical for the prevention of the disease. Optic disc and optic cup evaluation are one of the preeminent steps for glaucoma diagnosis. A novel approach is developed in this paper for the identification of glaucoma using a segmentation based approach on the optic disc and optic cup. The Dhristi dataset was used to help improve performance on a small dataset. A custom UNET++ model is built for the segmentation task by tuning the hyper-parameters in addition to a custom loss function. The developed loss function helps tackle the class imbalance occurring due to small size of the optic nerve head. The proposed approach achieves 96% accuracy in classifying glaucomatous and non-glaucomatous images based on clinical feature identification. The improvised model is able to achieve state-of-art results for Intersection over Union (IOU) scores, 0.9477 for optic disc and 0.9321 for optic cup, along with providing an enhancement in reducing the training time. The model was tested on publicly available datasets RIM-ONE, DRIONS-DB and ORIGA and is able to achieve an accuracy of 91%, 92% and 90% respectively. The developed approach is validated by training it over RIM-ONE dataset independently, without changing any model parameters. The model provides significant improvement in segmentation of the optic disc and optic cup along with improvement in training time.

Long-Term Follow-Up After Successful Trabeculectomy: A Case Report of Reversal of Cupping and Recovery of Visual Field Progression

Glaucoma is one of the leading causes of blindness worldwide, and reduction of intraocular pressure (IOP) is the only available evidence-based treatment that reduces visual field deterioration in glaucoma. We present a representative case of structural recovery and recovery of visual field progression after successful trabeculectomy (TLE) with long-term follow-up. A 35-year-old woman with glaucoma visited our hospital in 2008. The IOP in her right eye was 11 mmHg at the first visit, and subsequently increased to values in the high teens to 20 mmHg despite treatment with anti-glaucoma eye drops. During this period, the progression of this eye’s visual field deterioration was fast (mean deviation [MD] slope = −0.63 dB/year) and the optic disc cupping was advanced. In the seven-year period after successful TLE in 2014, the IOP declined to between 8 and 12 mmHg without any anti-glaucoma medication. During the first two years after TLE, the MD values were poorer than those before TLE. However, by 2020, MD values improved gradually to a state better than that before the surgery (MD slope during this period was +0.75 dB/year). The appearance of the optic disc was monitored using fundus photography, which showed optic disc morphological changes during topical glaucoma medication before and after TLE. In particular, a remarkable reversal of optic disc cupping enlargement started at two weeks after TLE, and the optic disc shape in 2021 was similar to that in 2008.Minimally invasive glaucoma surgeries are often preferred; however, our findings suggest that successful TLE followed by long-term rigorous IOP control can result in structural recoveries. Additionally, despite the deterioration in visual field in the early years after TLE, in the long term, it may settle down to the same level before the surgery with recovery of visual field progression, which may be a part of functional recovery.

A novel optic disc and optic cup segmentation technique to diagnose glaucoma using deep learning convolutional neural network over retinal fundus images

Glaucoma is currently leading retinal disease, which damages the eye because of the Intraocular pressure (IOP) on the eye. If glaucoma is left untreated it will lead to vision loss by damaging the Optic Nerve Head (ONH). The progression of glaucoma is examined on the retinal part of the eye by an experienced ophthalmologist. This approach is very tedious, and it consumes more time to do it manually. Hence this issue is right problem that can be solved by automatically diagnosing glaucoma with the help of the deep learning approaches. Convolutional Neural Networks (CNN's) are appropriate to find the solution for this type of issue as they can extract various levels of data from the input image, and which encourages to differentiate among non-glaucomic and glaucomic images. This proposed paper introduces an efficient glaucoma master framework to segment the optic cup and optic disc to find the Cup-to-Disc-Ratio (CDR). Here the diagnosis of glaucoma is achieved by using deep learning with novel CNN. The proposed system uses two individual CNN architecture to segment the Optic Cup (OC) and Optic Disc (OD) to get more accurate result. This model is trained and tested on DRISHTI – GS database, which is publicly available and an accuracy of 98% for optic disc and 97% for optic cup segmentation is achieved.

A multi-scale convolutional neural network with context for joint segmentation of optic disc and cup

Glaucoma is the leading cause of irreversible blindness. For glaucoma screening, the cup to disc ratio (CDR) is a significant indicator, whose calculation relies on the segmentation of optic disc(OD) and optic cup(OC) in color fundus images. This study proposes a residual multi-scale convolutional neural network with a context semantic extraction module to jointly segment the OD and OC. The proposed method uses a W-shaped backbone network, including image pyramid multi-scale input with the side output layer as an early classifier to generate local prediction output. The proposed method includes a context extraction module that extracts contextual semantic information from multiple level receptive field sizes and adaptively recalibrates channel-wise feature responses. It can effectively extract global information and reduce the semantic gaps in the fusion of deep and shallow semantic information. We validated the proposed method on four datasets, including DRISHTI-GS1, REFUGE, RIM-ONE r3, and a private dataset. The overlap errors are 0.0540, 0.0684, 0.0492, 0.0511 in OC segmentation and 0.2332, 0.1777, 0.2372, 0.2547 in OD segmentation, respectively. Experimental results indicate that the proposed method can estimate the CDR for a large-scale glaucoma screening.

GlaucoNet: Patch-Based Residual Deep Learning Network for Optic Disc and Cup Segmentation Towards Glaucoma Assessment

Glaucoma is a chronic eye condition causing irreversible vision damage and presently stands as the second leading cause of blindness worldwide. Damaged optic disc and optic cup assessment in color fundus image has been shown to be a promising method for glaucoma screening. In this paper, we propose a new accurate and efficient optic disc and cup segmentation methodology using a residual deep learning approach. The proposed patch-based deep network (GlaucoNet) is trained with a large number of preprocessed input image patches, big enough to include the important discriminatory information around each pixel. The proposed architecture gives better segmentation results as extra skip connections are introduced in the framework which explicitly reformulate the layers such that the learning function is dependent on a residual function of the input layer. It addresses the degradation problem of the deep network by improving the information flow. The convex hull transformation is applied on the initial results to obtain the final segmentation output. The proposed method is tested on publicly available DRISHTI-GS, RIM-ONE and ORIGA-light datasets. The experimental results for optic disc segmentation show an overlapping score (OS) of 0.9106, 0.8972 and 0.8835 and the optic cup segmentation achieves an OS of 0.8229, 0.7401 and 0.8106 in DRISHTI-GS, RIM-ONE and ORIGA-light, respectively. Finally, the glaucoma risk index is obtained by computing the cup-to-disc height ratio obtained from the segmented regions. Experimental results showing better or comparable segmentation performance and decrease in cup-to-disc height ratio estimation errors demonstrate the efficacy of the proposed GlaucoNet-based segmentation methodology.