Figure - available from: Journal of Ophthalmology
This content is subject to copyright. Terms and conditions apply.
Sample results of the optic cup. From left to right columns: (a) the original images, (b) the manual “ground truth,” and ((c)–(e)) outlines by the proposed method before ellipse fitting.
Source publication
![Optic disc in fundus image [14].](publication/284723118/figure/fig1/AS:1086777104306177@1636119261384/Optic-disc-in-fundus-image-14_Q320.jpg)
![The relation between the ground truth and automatically marked area [9].](publication/284723118/figure/fig4/AS:1086777104306188@1636119261521/The-relation-between-the-ground-truth-and-automatically-marked-area-9_Q320.jpg)
![Measurement of cup-to-disc ratio for a tilted disc [44].](publication/284723118/figure/fig5/AS:1086777104302098@1636119261573/Measurement-of-cup-to-disc-ratio-for-a-tilted-disc-44_Q320.jpg)
Glaucoma is the second leading cause of loss of vision in the world. Examining the head of optic nerve (cup-to-disc ratio) is very important for diagnosing glaucoma and for patient monitoring after diagnosis. Images of optic disc and optic cup are acquired by fundus camera as well as Optical Coherence Tomography. The optic disc and optic cup segmen...
Similar publications
![DIFFERENT DATABASES WITH IMAGE SIZE AND ESTIMATED SIZE FOR OD ROI [13].](profile/Amira-Sallow/publication/335575432/figure/tbl1/AS:917520856911873@1595765427874/DIFFERENT-DATABASES-WITH-IMAGE-SIZE-AND-ESTIMATED-SIZE-FOR-OD-ROI-13_Q320.jpg)
Segmentation of optical disk (OD) and blood vessel is one of the significant steps in automatic diabetic retinopathy (DR) detecting. In this paper, a new technique is presented for OD segmentation that depends on the histogram template matching algorithm and OD size. In addition, Kirsch method is used for Blood Vessel (BV) segmentation which is one...
Citations
... It used both unsupervised and supervised ML. Earlier, while detecting glaucomatous colour fundus photos, Glaucoma was classified by segmenting optic cup and as per primitive feature extraction ways [41][42][43]. ...
Glaucoma is a chronic progressive optic neuropathy characterized by damage to
the optic nerve and visual field loss. About 2.1 million people were found to have blindness by
developing Glaucoma worldwide. Glaucoma, after macular degeneration, is the second most
common reason for irreversible blindness 2.93% of people aged 40-80 years have Glaucoma,
among which many suffer open-angle Glaucoma. The evolution of Artificial Intelligence also
revolutionized the examination and treatment processes. AI has suitably fit itself in this
ophthalmic diagnosis. It is so because it highly magnifies even the deepest chambers of the
optic region, thereby creating ease for ophthalmologists.
... Noticeably, these features were applied as input feature vector to perform binary classification using SVM. Almazroa et al. [177] assessed the different segmentation models for cup and disk segmentation and allied boundary estimation for retinal abnormality detection. Gajbhiye and Kamthane [178] exploited geometric moment and wavelet features to perform glaucoma detection. ...
The incidence of diabetic retinopathy globally calls for advanced and more universally applicable computer-aided diagnosis (CAD) systems. This survey paper explores the current state of vision-based CAD techniques for the detection and classification of diabetic retinopathy, a diabetes-induced eye disorder that can lead to severe visual impairment or blindness. Characterized by a variety of manifestations including microaneurysms, exudates, hemorrhages, and macular detachment, diabetic retinopathy presents substantial challenges for automated detection. This is primarily due to the heterogeneity of retinal fundus images, which display diverse spatiotextural features and intricate vascular structures. Our exhaustive review indicates that most existing methodologies predominantly concentrate on isolated diabetic retinopathy types, employing localized spatiotextural feature analysis for classification. Such specificity often results in limited accuracy and generalizability, restricting practical real-world application. Furthermore, contemporary leading methods generally focus on single retinal characteristics, necessitating patients to undergo multiple CAD procedures, thereby increasing time, costs, and possibly intensifying retinal complexities. To overcome these obstacles, we propose the adoption of multi-trait-driven CAD solutions. Utilizing the potent capabilities of deep learning, these solutions could employ high-dimensional, multi-cue sensitive feature extraction and ensemble learning for classification. This approach is designed to improve the generalizability and dependability of CAD systems, offering a holistic solution capable of effectively managing the diverse manifestations of diabetic retinopathy. Our study highlights the need for a fundamental transformation in diabetic retinopathy CAD systems, motivating further research towards robust, multi-modal methods to enhance detection, classification, and grading of this widespread ailment.
... However, OCT is expensive, so many medical practitioners still use fundus images for diagnostic purposes [10]. Techniques like boundary detection, region segmentation, and color and contrast thresholding are widely used [11]. ...
... O diagnóstico do glaucomaé uma tarefa desafiadora e custosa, uma vez que depende de muitas observações feitas por oftalmologistas especialistas naárea. Com o avanço do uso de aprendizado profundo naárea de visão computacional, alguns trabalhos tiveram como objetivo utilizar esses modelos para auxiliar os oftalmologistas na detecção do glaucoma em imagens de fundo de olho [3] [4]. ...
... As pesquisas feitas podem ser divididas em dois tipos: As que focam no treinamento dos modelos para classificação binária das imagens entre a classe glaucoma e a classe normal, e as pesquisas focadas em utilizar métodos para segmentar á area do discoóptico e escavação [3] [4]. ...
... Devido ao fato de muitas doenças serem detectadas a partir de imagens médicas a aplicação de técnicas de aprendizado profundo podem ser muito benéficas, como por exemplo a automatização da análise da escavação e discoóptico pode auxiliar o oftalmologista no diagnóstico do glaucoma. Existem dois tipos de pesquisa que focam nesse objetivo: as que focam em classificação binária, e as que focam em classificações a partir de indicadores médicos como a relação disco/escavação e a regra ISNT [3]. ...
O glaucoma e uma doença silenciosa e progressiva que pode levar os pacientes a total perda de visão, por essa razão o diagnóstico precoce e importante. Métodos de visão computacional podem servir como um auxílio aos especialistas na detecção precoce. Esse trabalho teve como objetivo utilizar um modelo de rede neural artificial para detectar a área de interesse para detecção do glaucoma, e também fazer anotações da área de escavação e disco óptico. O modelo de detecção da área de interesse alcançou um mAP de 99,2%, e o modelo de detecção das áreas de escavação e disco óptico alcançou um mAP de 92%.
... It can take a long time to manually analyse retinal pictures [10]. To ensure the accuracy of the performed diagnosis, the images should also be examined by highly qualified ophthalmologists [11]. ...
One of the most serious eye illnesses, glaucoma affects the astrocytes and optic nerve fibres, causing irreversible damage to the eyes. As a result, glaucoma early identification is crucial in the medical industry. Retinal image-based detection falls within the category of non-invasive ways of detection among the many techniques. Automatic periodical screening can aid in the prompt detection of retinal glaucoma, while also easing the workload of skilled ophthalmologists. Effective glaucoma treatment can also lessen the severity of vision impairments brought on by the disease’s advanced stages. The retinal fundus dataset is used in this paper to undertake a novel glaucoma detection procedure. Additionally, the Scale-invariant feature transform (SIFT) is a widely used method for feature extraction and clustering in picture classification tasks. The characteristic is robust to variations in illumination, noise, partial occlusion, and minor changes in viewpoint in the photos. It is independent of the scale and orientation of the images. The real-time glaucoma screening system is suggested for use with this clustering approach. Convolutional neural network (CNN) classifier optimization is performed using a mix of the SIFT and Rooster Food Search optimization (RFSO) based algorithms for exploration and exploitation procedures. To assess whether a retina is glaucomatous or healthy, the resulting optic disc area is used. As a result, the study was presented based on efficient clustering, precise classification, and the use of various data sets, including the LAG and Rim-One database. The suggested technique was created in MATLAB and tested on this database. The suggested strategy produced accuracy levels above 95% and performed better on comparisons for additional metrics like clustering and classifier-related factors. In order to help researchers conduct additional study on glaucoma detection, we provide research difficulties and their respective solutions.
... Among these structures, the OD is considered the most important feature of retinal images, and it looks like a round disc structure with a bright appearance. In most algorithms for detecting retinal structures or diseases, the OD region can provide important prior information [2]. As mentioned above, in the fundus image, the OD is a relatively bright region. ...
Eye diseases often affect human health. Accurate detection of the optic disc contour is one of the important steps in diagnosing and treating eye diseases. However, the structure of fundus images is complex, and the optic disc region is often disturbed by blood vessels. Considering that the optic disc is usually a saliency region in fundus images, we propose a weakly-supervised optic disc detection method based on the fully convolution neural network (FCN) combined with the weighted low-rank matrix recovery model (WLRR). Firstly, we extract the low-level features of the fundus image and cluster the pixels using the Simple Linear Iterative Clustering (SLIC) algorithm to generate the feature matrix. Secondly, the top-down semantic prior information provided by FCN and bottom-up background prior information of the optic disc region are used to jointly construct the prior information weighting matrix, which more accurately guides the decomposition of the feature matrix into a sparse matrix representing the optic disc and a low-rank matrix representing the background. Experimental results on the DRISHTI-GS dataset and IDRiD dataset show that our method can segment the optic disc region accurately, and its performance is better than existing weakly-supervised optic disc segmentation methods. Graphical abstract of optic disc segmentation.
... Cupping or cup-to-disc ratio (CDR), which is the percentage of the OC's perpendicular diameter to the OD's perpendicular diameter, is an important metric for glaucoma diagnosis in computational approaches to the detection of glaucoma [17]. The most efficient technique to identify glaucoma is to examine the optic nerve head (ONH), which is separated into two regions: the optic disc (OD) and the optic cup (OC) [19], [20]. ...
Glaucoma is an eye disease that causes damage to the optic nerve, which can lead to visual loss and permanent blindness. Early glaucoma detection is therefore critical in order to avoid permanent blindness. The estimation of the cup-to-disc ratio (CDR) during an examination of the optical disc (OD) is used for the diagnosis of glaucoma. In this paper, we present the EDDense-Net segmentation network for the joint segmentation of OC and OD. The encoder and decoder in this network are made up of dense blocks with a grouped convolutional layer in each block, allowing the network to acquire and convey spatial information from the image while simultaneously reducing the network's complexity. To reduce spatial information loss, the optimal number of filters in all convolution layers were utilised. In semantic segmentation, dice pixel classification is employed in the decoder to alleviate the problem of class imbalance. The proposed network was evaluated on two publicly available datasets where it outperformed existing state-of-the-art methods in terms of accuracy and efficiency. For the diagnosis and analysis of glaucoma, this method can be used as a second opinion system to assist medical ophthalmologists.
... Around 1.2 million RGC axons converge to form the optic nerve head (ONH), also referred to as the optic papilla or optic disc. The ONH exhibits a brighter central depression known as the optic cup [51][52][53][54]. Blood supply to the ONH in humans is provided by the arterial circle of Zinn-Haller [48]. ...
Optic nerve disorders encompass a wide spectrum of conditions characterized by the loss of retinal ganglion cells (RGCs) and subsequent degeneration of the optic nerve. The etiology of these disorders can vary significantly, but emerging research highlights the crucial role of oxidative stress, an imbalance in the redox status characterized by an excess of reactive oxygen species (ROS), in driving cell death through apoptosis, autophagy, and inflammation. This review provides an overview of ROS-related processes underlying four extensively studied optic nerve diseases: glaucoma, Leber’s hereditary optic neuropathy (LHON), anterior ischemic optic neuropathy (AION), and optic neuritis (ON). Furthermore, we present preclinical findings on antioxidants, with the objective of evaluating the potential therapeutic benefits of targeting oxidative stress in the treatment of optic neuropathies.
... It used both unsupervised and supervised ML. Earlier, while detecting glaucomatous colour fundus photos, Glaucoma was classified by segmenting optic cup and as per primitive feature extraction ways [41][42][43]. ...
Background: Glaucoma is a chronic progressive optic neuropathy characterized by damage to
the optic nerve and visual field loss. About 2.1 million people were found to have blindness by
developing Glaucoma worldwide. Glaucoma, after macular degeneration, is the second most
common reason for irreversible blindness 2.93% of people aged 40-80 years have Glaucoma,
among which many suffer open-angle Glaucoma. The evolution of Artificial Intelligence also
revolutionized the examination and treatment processes. AI has suitably fit itself in this
ophthalmic diagnosis. It is so because it highly magnifies even the deepest chambers of the
optic region, thereby creating ease for ophthalmologists.
Aims: To analyze the efficacy of the selected programmed classifiers in the early detection of
Glaucoma for proper diagnosis and management at the early stage for better prognosis.
Methods: A cross-sectional study was conducted on healthy and Glaucoma patients. All the
patients underwent ophthalmic evaluation, and the tests were completed within six months.
The study selected seven machine learning classifiers trained for early diagnosis of Glaucoma
and to differentiate effectively from the normal eye. These classifiers were used appropriately
in the early diagnosis of Glaucoma. The study statistically analyzed the outcomes generated by
these classifiers.
Results: The study found an Area Under Curve (AUC) of 0.851, the worldwide vC/D, or
vertical cup/disc ratio, was at best. The global GPS has the highest AUC of the seven GPS
sectoral metrics (0.834). While not significantly better than global vC/D, RPART AUC in all
95 variables significantly improved over global GPS. Compared to both global vC/D & global
GPS, SVM-radial, including all 95 parameters, showed a significant improvement.
Conclusion: The study concluded that these classifiers are significantly efficient in
differentiating the glaucomatous eye from healthy ones at an early stage.
Keywords: Glaucoma, Classifiers, Artificial Intelligence, Eye, Photoreceptors.
... There exist several other parameters that are indicative risk factors of glaucoma, such as increased intraocular pressure [5], peripapillary atrophy (PPA) [32] or disc diameter [51]. However, since CDR is most commonly used, early detection and treatment heavily rely on the accurate detection of the optic disc and cup. ...
Glaucoma is one of the world leading causes of irreversible blindness. Early detection is essential to delay its progression and prevent vision loss. An accurate segmentation of the cup region in retinal fundus images is necessary to obtain relevant measurements for the detection of glaucoma. In recent years, multiple methods have been developed to automatically detect this region. All these methods are adjusted or trained using images that had been previously segmented by experts. In order to aid clinicians in performing this task, an interactive tool for the segmentation of the optic cup in stereo retinal fundus images using virtual reality has been developed. By using stereo images, the implemented virtual reality environment allows users to naturally perceive the three–dimensional structure of the optic cup region, which eases its segmentation compared to monocular images. The usage of the presented application was observed to increase accuracy of the delimitation, compared to using only two–dimensional fundus images, especially on areas with blood vessels.
