Fig 1 - available via license: Creative Commons Attribution-NonCommercial 4.0 International
Content may be subject to copyright.
Source publication
Background:
Existing methods may fail to locate and segment the optic disc (OD) due to imprecise boundaries, inconsistent image contrast and deceptive edge features in retinal images.
Objective:
To locate the OD and detect the OD boundary accurately.
Methods:
The method exploits a multi-stage strategy in the detection procedure. Firstly, OD lo...
Similar publications
The stability and quantity of feature matching in video sequence is one of the key issues for feature tracking and some relevant applications. The existing matching methods are based on feature detection, which is usually affected by illumination conditions, noise or occlusions, and this will directly influence matching results. In this paper, we p...
Citations
... Accurate OD segmentation in fundus images is a challenging task for the following reasons ( Figure 3): (1) Some parts of OD boundaries are covered by blood vessels; (2) some types of noise such as parapapillary atrophy (PPA) and bright area affect OD segmentation. For Reason (1), blood vessels can be removed by using a mathematical morphology [16]; however, some bright noise and PPA are mixed in the OD area, resulting in undersegmentation. For Reason (2), according to the different textures between PPA and other regions, a gray level co-occurrence matrix (GLCM) can be used to detect the PPA [17]. ...
... Proposed OD segmentation algorithms can be roughly divided into five categories: threshold- [18,19], pattern- [20][21][22][23], classification and clustering- [24][25][26], active contour model- [16,[27][28][29][30][31][32][33], and deep learning- [34][35][36]-based methods. In threshold-based methods, the ODwas segmented by OTSU and other thresholding methods in [18,19]. ...
... However, segmentation fails if the OD boundary is extremely smooth. The level set model was improved in [16,28] by applying multiple energy functionals that included gradient-, area-, and shape-based energy functionals. A shape-based functional is able to limit the shape of the contour, the area-based functional is adopted to change the position and size of the prior shape model, and the gradient-based functional is the main source of energy to find the OD boundaries. ...
The accurate segmentation of the optic disc (OD) in fundus images is a crucial step for the analysis of many retinal diseases. However, because of problems such as vascular occlusion, parapapillary atrophy (PPA), and low contrast, accurate OD segmentation is still a challenging task. Therefore, this paper proposes a multiple preprocessing hybrid level set model (HLSM) based on area and shape for OD segmentation. The area-based term represents the difference of average pixel values between the inside and outside of a contour, while the shape-based term measures the distance between a prior shape model and the contour. The average intersection over union (IoU) of the proposed method was 0.9275, and the average four-side evaluation (FSE) was 4.6426 on a public dataset with narrow-angle fundus images. The IoU was 0.8179 and the average FSE was 3.5946 on a wide-angle fundus image dataset compiled from a hospital. The results indicate that the proposed multiple preprocessing HLSM is effective in OD segmentation.
... In the previous studies, the OD was segmented by using handcrafted features, including image gradients [22], stereo image features [23], local texture features [24], and superpixel based classification [25]. A level-set based automatic algorithm [26]was proposed for OD localization and segmentation. Compared with OD, the OC is harder to segment due to its low contrast boundary and the vascular occlusions. ...
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.
... Apart from this, top hat and thresholding was applied to extract the vessels and create its mask for removing the vessels region completely. Finally, Gaussian filter was used to smoothen the processed image for further analysis [49]. Ayub et al. started with extraction of ROI based on highest intensity of optic disc centre. ...
Early diagnosis of diseases related with retina such as glaucoma is of utmost importance in current scenario as it is the second most prevailing cause of irreversible blindness over the world and is expected to increase further in near future. It is commonly diagnosed using retinal images which are acquired by digital fundus cameras. But the acquired images may be prone to certain outliers that create hindrance in diagnosis of glaucoma by tempering the accuracy. These outliers include retinal vessels, low contrast of images and uneven illumination that deteriorates the performance of disc and cup segmentation which are the key indicators to diagnose glaucoma. Thus, pre-processing of retinal images to remove outliers plays a significant role in diagnosis. This paper presents an approach for pre-processing the retinal fundus image followed by its comparison with state of the art. Based on the experimental analysis the performance of the proposed approach is found to be better than the state of the art based on the analysis using metrics such as peak signal to ratio, mean square error and structural similarity index. Further, the proposed approach has been compared with state of the art using metrics such as Jaccard index and dice similarity on the basis of segmentation outcomes on different pre-processing approaches.
... In addition, some lesions (such as hard exudation) with similar brightness and shape as the optic disc will also render this method unsuccessful. The third method is based on the active contour model, which first determines the initial contour of the optic disc and then obtains the boundary of the optic disc by using the continuous evolution of the contour driven by external constraints and internal energy (29). ...
The emergence of computer graphics processing units (GPUs), improvements in mathematical models, and the availability of big data, has allowed artificial intelligence (AI) to use machine learning and deep learning (DL) technology to achieve robust performance in various fields of medicine. The DL system provides improved capabilities, especially in image recognition and image processing. Recent progress in the sorting of AI data sets has stimulated great interest in the development of DL algorithms. Compared with subjective evaluation and other traditional methods, DL algorithms can identify diseases faster and more accurately in diagnostic tests. Medical imaging is of great significance in the clinical diagnosis and individualized treatment of ophthalmic diseases. Based on the morphological data sets of millions of data points, various image-related diagnostic techniques can now impart high-resolution information on anatomical and functional changes, thereby providing unprecedented insights in ophthalmic clinical practice. As ophthalmology relies heavily on imaging examinations, it is one of the first medical fields to apply DL algorithms in clinical practice. Such algorithms can assist in the analysis of large amounts of data acquired from the examination of auxiliary images. In recent years, rapid advancements in imaging technology have facilitated the application of DL in the automatic identification and classification of pathologies that are characteristic of ophthalmic diseases, thereby providing high quality diagnostic information. This paper reviews the origins, development, and application of DL technology. The technical and clinical problems associated with building DL systems to meet clinical needs and the potential challenges of clinical application are discussed, especially in relation to the field of optic nerve diseases.
... In the previous studies, the OD was segmented by using handcrafted features, including image gradients [22], stereo image features [23], local texture features [24], and superpixel based classification [25]. A level-set based automatic algorithm [26]was proposed for OD localization and segmentation. Compared with OD, the OC is harder to segment due to its low contrast boundary and the vascular occlusions. ...
Glaucoma is the leading cause of irreversible blindness, but its early symptoms are not obvious and are easily overlooked, so early screening for glaucoma is particularly important. The cup to disc ratio is an important indicator for clinical glaucoma screening, and accurate segmentation of the optic cup and disc is the key to calculating the cup to disc ratio. In this paper, a full convolutional neural network with residual multi-scale convolution module was proposed for the optic cup and disc segmentation. First, the fundus image was contrast enhanced and polar transformation was introduced. Subsequently, W-Net was used as the backbone network, which replaced the standard convolution unit with the residual multi-scale full convolution module, the input port was added to the image pyramid to construct the multi-scale input, and the side output layer was used as the early classifier to generate the local prediction output. Finally, a new multi-tag loss function was proposed to guide network segmentation. The mean intersection over union of the optic cup and disc segmentation in the REFUGE dataset was 0.904 0 and 0.955 3 respectively, and the overlapping error was 0.178 0 and 0.066 5 respectively. The results show that this method not only realizes the joint segmentation of cup and disc, but also improves the segmentation accuracy effectively, which could be helpful for the promotion of large-scale early glaucoma screening.
... In the previous studies, the OD was segmented by using handcrafted features, including image gradients [22], stereo image features [23], local texture features [24], and superpixel based classification [25]. A level-set based automatic algorithm [26]was proposed for OD localization and segmentation. Compared with OD, the OC is harder to segment due to its low contrast boundary and the vascular occlusions. ...
Laryngeal leukoplakia is one kind of precancerous lesions in the larynx. Precise detection and segmentation of leukoplakia in laryngoscopic images is important for laryngeal disease diagnosis and treatment. In this paper, we proposed a multi-scale recurrent fully convolution neural network named boldface-M-Net (BM-Net) to identify and segment laryngeal leukoplakia lesions. The proposed BM-Net was composed of a multi-scale input layer, a double U-shaped convolution network, and a side-output layer. First, we augmented the image to produce six channels and then constructed image pyramids for the multi-scale input layer. For the U-shaped convolution network, we constructed a new U-Net using multi-scale convolution and a recurrent convolution layer (RCL) instead of the original convolution layer. We then employed skip connections to connect the double U-shaped convolution network, one with three 2 × 2 max pooling layers and the other with four, thus forming the main structure of BM-Net. We added the output for the three-layered U-Net to the side-output layer to produce a companion local prediction map for each scale layer. Image pyramids and multi-scale convolution can generate multiple level-receptive fields, while the RCL allows for the greater perception of context with parameter t increases. Finally, we compared the performance of the proposed BM-Net with the popular networks, including FCN-8s, Seg-Net, U-Net, M-Net, and three other modified networks for segmenting laryngeal leukoplakia in laryngoscopic images. According to the experimental results, BM-Net, which inherited the advantages of U-Net, M-Net, and RCL, exhibited overall better performance in laryngeal leukoplakia segmentation than the other networks.
... In the same study, by applying the Circular Hough Transformation on this extracted image, they were able to detect the OD location, and a 97% success rate was achieved over 40 images. The literature shows that hybrid methods have been proposed for automatic segmentation of the OD [10,11,12]. For example, In [10] was developed a fully automated method using the region growing method and the L1 minimization algorithm, which are semi-automatic methods for obtaining OD segmentation. ...
... In [11], authors segmented the OD using the hybrid vessel phase portrait analysis, where the results showed that true negatives and true positives were obtained in 94.67 and 98.13% of cases, respectively. In [12], OD was segmented with 98.67% success using a hybrid approach involving the ACM and ellipse fit method. In [13], OD was determined by histogram matching technique to identify the presence of pathological regions. ...
... TP=98.13% [12] Active Contour, Ellipse Fit Method A=98.67% [13] Histogram Matching A=100% A=98.9% A=91.36% [14] Circular Transformation A=98.77% A=97.5% A= 99.75% [15] Brighness Based Method, Supervised Descent Method O=89.44% [16] Sparse Different applied methods to OD segmentation were presented in Table.2. ...
This paper proposes a hybrid method that is capable of automatically implementing the Optic Disc (OD) segmentation. In the hybrid method, two steps were performed. First, the location of the OD was determined by template matching. Second, the OD location coordinates obtained in the first stage were given as inputs to the Active Contour Method applied to complete the OD segmentation. Furthermore, as part of this study, Android based a program was developed to allow physicians the ability to independently access the proposed hybrid method results from anywhere and to add comments. Thus, the physician would be able to instantly track the patient. Performance evaluation of the proposed hybrid method was done separately for both localization and segmentation. The success of localization was confirmed on the basis of whether the determined coordinates corresponded to the OD. The segmentation process was assessed according to the parameters, as derived from a confusion matrix. The average Dice coefficient obtained for all images was 0.943, while the average values of accuracy, specificity and sensitivity parameters for all images were calculated as 0.90, 0.961 and 0.931, respectively. The final results obtained from the proposed hybrid method were checked by a physician, who observed that the OD was successfully segmented.
... Loss of these axons underpins the pattern of field loss observed in glaucoma. The OD typically presents on a fundus image as a bright circle or ellipse with yellow or white coloring [39] and is usually identified from the red channel of an image due to the highest contrast difference with the background retina [40,41]. The location of the OD can be used as a reference point to establish the position of the macula on the fundus image [42]. ...
Introduction: Retinal imaging is a fundamental tool in ophthalmic diagnostics. The potential use of retinal imaging within screening programs, with consequent need to analyze large numbers of images with high throughput, is pushing the digital image analysis field to find new solutions for the extraction of specific information from the retinal image. Areas covered: The aim of this review is to explore the latest progress in image processing techniques able to recognize specific retinal image features and potential features of disease. In particular, this review aims to describe publically available retinal image databases, highlight different performance evaluators commonly used within the field, outline current approaches in feature-based retinal image analysis and to map related trends. Expert Commentary: This review found two key areas to be addressed for the future development of automatic retinal image analysis: fundus image quality and the affect image processing may impose on relevant clinical information within the images. Performance evaluators of the algorithms reviewed are very promising, however absolute values are difficult to interpret when validating system suitability for use within clinical practice.
Background:
Diabetic macular edema (DME) is one of the severe complication of diabetic retinopathy causing severe vision loss and leads to blindness in severe cases if left untreated.
Objective:
To grade the severity of DME in retinal images.
Methods:
Firstly, the macular is localized using its anatomical features and the information of the macula location with respect to the optic disc. Secondly, a novel method for the exudates detection is proposed. The possible exudate regions are segmented using vector quantization technique and formulated using a set of feature vectors. A semi-supervised learning with graph based classifier is employed to identify the true exudates. Thirdly, the disease severity is graded into different stages based on the location of exudates and the macula coordinates.
Results:
The results are obtained with the mean value of 0.975 and 0.942 for accuracy and F1-scrore, respectively.
Conclusion:
The present work contributes to macula localization, exudate candidate identification with vector quantization and exudate candidate classification with semi-supervised learning. The proposed method and the state-of-the-art approaches are compared in terms of performance, and experimental results show the proposed system overcomes the challenge of the DME grading and demonstrate a promising effectiveness.
