Accurate and efficient optic disc detection and segmentation by a circular transformation

W-Net: A boundary-aware cascade network for robust and accurate optic disc segmentation

Accurate optic disc (OD) segmentation has a great significance for computer-aided diagnosis of different types of eye diseases. Due to differences in image acquisition equipment and acquisition methods, the resolution, size, contrast, and clarity of images from different datasets show significant differences, resulting in poor generalization performance of deep learning networks. To solve this problem, this study proposes a multi-level segmentation network. The network includes data quality enhancement module (DQEM), coarse segmentation module (CSM), localization module (OLM), and fine segmentation stage module (FSM). In FSM, W-Net is proposed for the first time, and boundary loss is introduced in the loss function, which effectively improves the performance of OD segmentation. We generalized the model in the REFUGE test dataset, GAMMA dataset, Drishti-GS1 dataset, and IDRiD dataset, respectively. The results show that our method has the best OD segmentation performance in different datasets compared with state-of-the-art networks.

Superpixel-guided class-level denoising for unsupervised domain adaptive fundus image segmentation without source data

Unsupervised domain adaptation (UDA), which is used to alleviate the domain shift between the source domain and target domain, has attracted substantial research interest. Previous studies have proposed effective UDA methods which require both labeled source data and unlabeled target data to achieve desirable distribution alignment. However, due to privacy concerns, the vendor side often can only trade the pretrained source model without providing the source data to the targeted client, leading to failed adaptation by classical UDA techniques. To address this issue, in this paper, a novel Superpixel-guided Class-level Denoised self-training framework (SCD) is proposed, aiming at effectively adapting the pretrained source model to the target domain in the absence of source data. Since the source data is unavailable, the model can only be trained on the target domain with the pseudo labels obtained from the pretrained source model. However, due to domain shift, the predictions obtained by the source model on the target domain are noisy. Considering this, we propose three mutual-reinforcing components tailored to our self-training framework: (i) an adaptive class-aware thresholding strategy for more balanced pseudo label generation, (ii) a masked superpixel-guided clustering method for generating multiple content-adaptive and spatial-adaptive feature centroids that enhance the discriminability of final prototypes for effective prototypical label denoising, and (iii) adaptive learning schemes for suspected noisy-labeled and correct-labeled pixels to effectively utilize the valuable information available. Comprehensive experiments on multi-site fundus image segmentation demonstrate the superior performance of our approach and the effectiveness of each component.

ECLNet: Center localization of eye structures based on Adaptive Gaussian Ellipse Heatmap

Accurately localizing the center of specific biological structures in medical images is of great significance for clinical treatment. The center localization task can be viewed as an estimation problem of keypoints, and the heatmap is often used to describe the probability of the location of keypoints during estimation. Existing methods construct the heatmap from a Gaussian kernel function with a fixed standard deviation, therefore cannot adapt to morphologic changes of the target region. In this paper, we build a deep network, ECLNet, to localize the center of eye-related structures in medical images. Meanwhile, we propose a method called Adaptive Gaussian Ellipse Heatmap (AGEH), which can efficiently utilize the gradient feature of the target region to adjust the morphology of the heatmap. The ECLNet localizes the optic disc and fovea center with mean Euclidean Distance of 17.995 and 39.446 pixels, respectively, for IDRiD dataset. The ECLNet also successfully localizes the eye center with the mean absolute Position Error of 0.186±0.027 mm for CATARACT dataset. The results show that our proposed method has a better performance compared with some state-of-the-art methods.

Which Color Channel Is Better for Diagnosing Retinal Diseases Automatically in Color Fundus Photographs?

Color fundus photographs are the most common type of image used for automatic diagnosis of retinal diseases and abnormalities. As all color photographs, these images contain information about three primary colors, i.e., red, green, and blue, in three separate color channels. This work aims to understand the impact of each channel in the automatic diagnosis of retinal diseases and abnormalities. To this end, the existing works are surveyed extensively to explore which color channel is used most commonly for automatically detecting four leading causes of blindness and one retinal abnormality along with segmenting three retinal landmarks. From this survey, it is clear that all channels together are typically used for neural network-based systems, whereas for non-neural network-based systems, the green channel is most commonly used. However, from the previous works, no conclusion can be drawn regarding the importance of the different channels. Therefore, systematic experiments are conducted to analyse this. A well-known U-shaped deep neural network (U-Net) is used to investigate which color channel is best for segmenting one retinal abnormality and three retinal landmarks.

Vessel-based hybrid optic disk segmentation applied to mobile phone camera retinal images

Precise detection of the optic disk (OD) is an important task in the diagnosis of diabetic retinopathy. To manage the massive diabetic population, there is a significant demand for efficient and remote retinal imaging techniques. In this regard, the use of handheld mobile cameras attached to a smartphone is a promising approach. However, smartphone retinal images are often of low quality, compared to those obtained on standard equipment. They also have a narrow field of view and an incomplete/unbalanced vessel structure. Hence, we propose a new, fully automatic hybrid method for OD localization (HLM). It is designed for and verified on mobile camera/smartphone retinal images. The HLM analyzes the vessel structure and finds the OD locations by using the exclusion method when an image has a complete vessel system, and a newly proposed line detection method, otherwise. For OD segmentation, an active contour model followed by the circle fitting approach is integrated into the HLM. The proposed method was tested on three mobile camera datasets and four datasets obtained by standard equipment. For mobile camera datasets, the HLM achieves an average accuracy of 98% for OD localization. The segmentation routine obtains an average precision of 92.64% and an average recall of 82.38%. Testing against the recent state-of-the-art methods on the standard datasets shows comparable performance. The proposed framework for OD localization and segmentation designed for and verified on mobile camera retinal datasets and standard datasets. (EM - "Exclusion Method", LDM - "Line Detection Method", OD - "Optic Disk" and PPV - "Positive Predictive Value").

A Novel Hybrid Approach Based on Deep CNN to Detect Glaucoma Using Fundus Imaging

Glaucoma is one of the eye diseases stimulated by the fluid pressure that increases in the eyes, damaging the optic nerves and causing partial or complete vision loss. As Glaucoma appears in later stages and it is a slow disease, detailed screening and detection of the retinal images is required to avoid vision forfeiture. This study aims to detect glaucoma at early stages with the help of deep learning-based feature extraction. Retinal fundus images are utilized for the training and testing of our proposed model. In the first step, images are pre-processed, before the region of interest (ROI) is extracted employing segmentation. Then, features of the optic disc (OD) are extracted from the images containing optic cup (OC) utilizing the hybrid features descriptors, i.e., convolutional neural network (CNN), local binary patterns (LBP), histogram of oriented gradients (HOG), and speeded up robust features (SURF). Moreover, low-level features are extracted using HOG, whereas texture features are extracted using the LBP and SURF descriptors. Furthermore, high-level features are computed using CNN. Additionally, we have employed a feature selection and ranking technique, i.e., the MR-MR method, to select the most representative features. In the end, multi-class classifiers, i.e., support vector machine (SVM), random forest (RF), and K-nearest neighbor (KNN), are employed for the classification of fundus images as healthy or diseased. To assess the performance of the proposed system, various experiments have been performed using combinations of the aforementioned algorithms that show the proposed model based on the RF algorithm with HOG, CNN, LBP, and SURF feature descriptors, providing ≤99% accuracy on benchmark datasets and 98.8% on k-fold cross-validation for the early detection of glaucoma.

Unsupervised Domain Adaptation Based Image Synthesis and Feature Alignment for Joint Optic Disc and Cup Segmentation

Due to the discrepancy of different devices for fundus image collection, a well-trained neural network is usually unsuitable for another new dataset. To solve this problem, the unsupervised domain adaptation strategy attracts a lot of attentions. In this paper, we propose an unsupervised domain adaptation method based image synthesis and feature alignment (ISFA) method to segment optic disc and cup on the fundus image. The GAN-based image synthesis (IS) mechanism along with the boundary information of optic disc and cup is utilized to generate target-like query images, which serves as the intermediate latent space between source domain and target domain images to alleviate the domain shift problem. Specifically, we use content and style feature alignment (CSFA) to ensure the feature consistency among source domain images, target-like query images and target domain images. The adversarial learning is used to extract domain invariant features for output-level feature alignment (OLFA). To enhance the representation ability of domain-invariant boundary structure information, we introduce the edge attention module (EAM) for low-level feature maps. Eventually, we train our proposed method on the training set of the REFUGE challenge dataset and test it on Drishti-GS and RIM-ONE_r3 datasets. On the Drishti-GS dataset, our method achieves about 3% improvement of Dice on optic cup segmentation over the next best method. We comprehensively discuss the robustness of our method for small dataset domain adaptation. The experimental results also demonstrate the effectiveness of our method. Our code is available at https://github.com/thinkobj/ISFA.

A robust method for the automatic location of the optic disc and the fovea in fundus images

BACKGROUND AND OBJECTIVE

The location of the optic disc (OD) and the fovea is usually crucial in automatic screening systems for diabetic retinopathy. Previous methods aimed at their location often fail when these structures do not have the standard appearance. The purpose of this work is to propose novel, robust methods for the automatic detection of the OD and the fovea.

METHODS

The proposed method comprises a preprocessing stage, a method for retinal background extraction, a vasculature segmentation phase and the computation of various novel saliency maps. The main novelty of this work is the combination of the proposed saliency maps, which represent the spatial relationships between some structures of the retina and the visual appearance of the OD and fovea. Another contribution is the method to extract the retinal background, based on region-growing.

RESULTS

The proposed methods were evaluated over a proprietary database and three public databases: DRIVE, DiaretDB1 and Messidor. For the OD, we achieved 100% accuracy for all databases except Messidor (99.50%). As for the fovea location, we also reached 100% accuracy for all databases except Messidor (99.67%).

CONCLUSIONS

Our results suggest that the proposed methods are robust and effective to automatically detect the OD and the fovea. This way, they can be useful in automatic screening systems for diabetic retinopathy as well as other retinal diseases.

A region growing and local adaptive thresholding-based optic disc detection

Automatic optic disc (OD) localization and segmentation is not a simple process as the OD appearance and size may significantly vary from person to person. This paper presents a novel approach for OD localization and segmentation which is fast as well as robust. In the proposed method, the image is first enhanced by de-hazing and then cropped around the OD region. The cropped image is converted to HSV domain and then V channel is used for OD detection. The vessels are extracted from the Green channel in the cropped region by multi-scale line detector and then removed by the Laplace Transform. Local adaptive thresholding and region growing are applied for binarization. Furthermore, two region properties, eccentricity, and area are then used to detect the true OD region. Finally, ellipse fitting is used to fill the region. Several datasets are used for testing the proposed method. Test results show that the accuracy and sensitivity of the proposed method are much higher than the existing state-of-the-art methods.

Automatic detection of parapapillary atrophy and its association with children myopia

BACKGROUND AND OBJECTIVE

To develop an automatic parapapillary atrophy (PPA) detection algorithm in retinal fundus images and discuss the association between PPA and myopia to facilitate diagnosis and prediction of children myopia.

METHODS

The proposed algorithm consists of PPA identification and segmentation, which are evaluated by comparing with ophthalmologist's annotation. The association between PPA parameters and myopia is analyzed via Spearman correlation.

RESULTS

The accuracy of PPA identification reaches 90.78%. The F1-score of PPA segmentation is 0.67, and the Pearson correlation between the automatic measurement and ground truths for the area of PPA (A_PPA), the ratio (μ) of A_PPA to the area of optic disc (OD) and the maximal width of PPA (W) are 0.74, 0.60, and 0.69 (all p < 0.001). All these parameter changes are significantly correlated with the change of ratio of axial length to corneal curvature (ΔALCC), spherical equivalent (ΔSE), and axial length (ΔAL) (all p < 0.01), in which the highest association is 0.75 between ΔW (the change of W) and ΔALCC.

CONCLUSIONS

The proposed algorithm can provide accurate PPA measurement. Strong association between the changes of PPA and the progress of children myopia are observed and the width of PPA has the best association among three PPA parameters.

A coarse-to-fine deep learning framework for optic disc segmentation in fundus images

Accurate segmentation of the optic disc (OD) depicted on color fundus images may aid in the early detection and quantitative diagnosis of retinal diseases, such as glaucoma and optic atrophy. In this study, we proposed a coarse-to-fine deep learning framework on the basis of a classical convolutional neural network (CNN), known as the U-net model, to accurately identify the optic disc. This network was trained separately on color fundus images and their grayscale vessel density maps, leading to two different segmentation results from the entire image. We combined the results using an overlap strategy to identify a local image patch (disc candidate region), which was then fed into the U-net model for further segmentation. Our experiments demonstrated that the developed framework achieved an average intersection over union (IoU) and a dice similarity coefficient (DSC) of 89.1% and 93.9%, respectively, based on 2,978 test images from our collected dataset and six public datasets, as compared to 87.4% and 92.5% obtained by only using the sole U-net model. The comparison with available approaches demonstrated a reliable and relatively high performance of the proposed deep learning framework in automated OD segmentation.

Region-segmentation strategy for Bruch's membrane opening detection in spectral domain optical coherence tomography images

Bruch's membrane opening (BMO) is an important biomarker in the progression of glaucoma. Bruch's membrane opening minimum rim width (BMO-MRW), cup-to-disc ratio in spectral domain optical coherence tomography (SD-OCT) and lamina cribrosa depth based on BMO are important measurable parameters for glaucoma diagnosis. The accuracy of measuring these parameters is significantly affected by BMO detection. In this paper, we propose a method for automatically detecting BMO in SD-OCT volumes accurately to reduce the impact of the border tissue and vessel shadows. The method includes three stages: a coarse detection stage composed by retinal pigment epithelium layer segmentation, optic disc segmentation, and multi-modal registration; a fixed detection stage based on the U-net in which BMO detection is transformed into a region segmentation problem and an area bias component is proposed in the loss function; and a post-processing stage based on the consistency of results to remove outliers. Experimental results show that the proposed method outperforms previous methods and achieves a mean error of 42.38 μm.

Diabetic retinopathy techniques in retinal images: A review

The diabetic retinopathy is the main reason of vision loss in people. Medical experts recognize some clinical, geometrical and haemodynamic features of diabetic retinopathy. These features include the blood vessel area, exudates, microaneurysm, hemorrhages and neovascularization, etc. In Computer Aided Diagnosis (CAD) systems, these features are detected in fundus images using computer vision techniques. In this paper, we review the methods of low, middle and high level vision for automatic detection and classification of diabetic retinopathy.We give a detailed review of 79 algorithms for detecting different features of diabetic retinopathy during the last eight years.

The region of interest localization for glaucoma analysis from retinal fundus image using deep learning

BACKGROUND AND OBJECTIVES

Retinal fundus image analysis without manual intervention has been rising as an imperative analytical approach for early detection of eye-related diseases such as glaucoma and diabetic retinopathy. For analysis and detection of Glaucoma and some other disease from retinal image, there is a significant role of predicting the bounding box coordinates of Optic Disc (OD) that acts as a Region of Interest (ROI).

METHODS

We reframe ROI detection as a solitary regression predicament, from image pixel values to ROI coordinates including class probabilities. A Convolution Neural Network (CNN) has trained on full images to predict bounding boxes along with their analogous probabilities and confidence scores. The publically available MESSIDOR and Kaggle datasets have been used to train the network. We adopted various data augmentation techniques to amplify our dataset so that our network becomes less sensitive to noise. From a very high-level perspective, every image is divided into a 13 × 13 grid. Every grid cell envisages 5 bounding boxes along with the corresponding class probability and a confidence score. Before training, the network and the bounding box priors or anchors are initialized using k-means clustering on the original dataset using a distance metric based on Intersection of the Union (IOU) over ground-truth bounding boxes. During training in fact, a sum-squared loss function is used as the prediction's error function. Finally, Non-maximum suppression is applied by the proposed methodology to reach the concluding prediction.

RESULTS

The following projected method accomplish an accuracy of 99.05% and 98.78% on the Kaggle and MESSIDOR test sets for ROI detection. Results of proposed methodology indicates that proposed network is able to perceive ROI in fundus images in 0.0045 s at 25 ms of latency, which is far better than the recent-time and using no handcrafted features.

CONCLUSIONS

The network predicts accurate results even on low-quality images without being biased towards any particular type of image. The network prepared to see more summed up depiction rather than past works in the field. Going by the results, our novel method has better diagnosis of eye diseases in the future in a faster and reliable way.

Joint Optic Disc and Cup Segmentation Based on Multi-Label Deep Network and Polar Transformation

Glaucoma is a chronic eye disease that leads to irreversible vision loss. The cup to disc ratio (CDR) plays an important role in the screening and diagnosis of glaucoma. Thus, the accurate and automatic segmentation of optic disc (OD) and optic cup (OC) from fundus images is a fundamental task. Most existing methods segment them separately, and rely on hand-crafted visual feature from fundus images. In this paper, we propose a deep learning architecture, named M-Net, which solves the OD and OC segmentation jointly in a one-stage multi-label system. The proposed M-Net mainly consists of multi-scale input layer, U-shape convolutional network, side-output layer, and multi-label loss function. The multi-scale input layer constructs an image pyramid to achieve multiple level receptive field sizes. The U-shape convolutional network is employed as the main body network structure to learn the rich hierarchical representation, while the side-output layer acts as an early classifier that produces a companion local prediction map for different scale layers. Finally, a multi-label loss function is proposed to generate the final segmentation map. For improving the segmentation performance further, we also introduce the polar transformation, which provides the representation of the original image in the polar coordinate system. The experiments show that our M-Net system achieves state-of-the-art OD and OC segmentation result on ORIGA data set. Simultaneously, the proposed method also obtains the satisfactory glaucoma screening performances with calculated CDR value on both ORIGA and SCES datasets.

Mean curvature and texture constrained composite weighted random walk algorithm for optic disc segmentation towards glaucoma screening

Accurate optic disc (OD) segmentation is an important step in obtaining cup-to-disc ratio-based glaucoma screening using fundus imaging. It is a challenging task because of the subtle OD boundary, blood vessel occlusion and intensity inhomogeneity. In this Letter, the authors propose an improved version of the random walk algorithm for OD segmentation to tackle such challenges. The algorithm incorporates the mean curvature and Gabor texture energy features to define the new composite weight function to compute the edge weights. Unlike the deformable model-based OD segmentation techniques, the proposed algorithm remains unaffected by curve initialisation and local energy minima problem. The effectiveness of the proposed method is verified with DRIVE, DIARETDB1, DRISHTI-GS and MESSIDOR database images using the performance measures such as mean absolute distance, overlapping ratio, dice coefficient, sensitivity, specificity and precision. The obtained OD segmentation results and quantitative performance measures show robustness and superiority of the proposed algorithm in handling the complex challenges in OD segmentation.

Template Matching via Densities on the Roto-Translation Group

We propose a template matching method for the detection of 2D image objects that are characterized by orientation patterns. Our method is based on data representations via orientation scores, which are functions on the space of positions and orientations, and which are obtained via a wavelet-type transform. This new representation allows us to detect orientation patterns in an intuitive and direct way, namely via cross-correlations. Additionally, we propose a generalized linear regression framework for the construction of suitable templates using smoothing splines. Here, it is important to recognize a curved geometry on the position-orientation domain, which we identify with the Lie group SE(2): the roto-translation group. Templates are then optimized in a B-spline basis, and smoothness is defined with respect to the curved geometry. We achieve state-of-the-art results on three different applications: detection of the optic nerve head in the retina (99.83 percent success rate on 1,737 images), of the fovea in the retina (99.32 percent success rate on 1,616 images), and of the pupil in regular camera images (95.86 percent on 1,521 images). The high performance is due to inclusion of both intensity and orientation features with effective geometric priors in the template matching. Moreover, our method is fast due to a cross-correlation based matching approach.

Contrast based circular approximation for accurate and robust optic disc segmentation in retinal images

A new method for automatic optic disc localization and segmentation is presented. The localization procedure combines vascular and brightness information to provide the best estimate of the optic disc center which is the starting point for the segmentation algorithm. A detection rate of 99.58% and 100% was achieved for the Messidor and ONHSD databases, respectively. A simple circular approximation to the optic disc boundary is proposed based on the maximum average contrast between the inner and outer ring of a circle centered on the estimated location. An average overlap coefficient of 0.890 and 0.865 was achieved for the same datasets, outperforming other state of the art methods. The results obtained confirm the advantages of using a simple circular model under non-ideal conditions as opposed to more complex deformable models.

Automatic optic disk detection in retinal images using hybrid vessel phase portrait analysis

We propose vessel vector-based phase portrait analysis (VVPPA) and a hybrid between VVPPA and a clustering method proposed earlier for automatic optic disk (OD) detection called the vessel transform (VT). The algorithms are based primarily on the location and direction of retinal blood vessels and work equally well on fine and poor quality images. To localize the OD, the direction vectors derived from the vessel network are constructed, and points of convergence of the resulting vector field are examined by phase portrait analysis. The hybrid method (HM) uses a set of rules acquired from the decision model to alternate the use of VVPPA and VT. To identify the OD contour, the scale space (SS) approach is integrated with VVPPA, HM, and the circular approximation (SSVVPPAC and SSHMC). We test the proposed combination against state-of-the-art OD detection methods. The results show that the proposed algorithms outperform the benchmark methods, especially on poor quality images. Specifically, the HM gets the highest accuracy of 98% for localization of the OD regardless of the image quality. Testing the segmentation routines SSVVPPAC and SSHMC against the conventional methods shows that SSHMC performs better than the existing methods, achieving the highest PPV of 71.81% and the highest sensitivity of 70.67% for poor quality images. Furthermore, the HM can supplement practically any segmentation model as long as it offers multiple OD candidates. In order to prove this claim, we test the efficiency of the HM in detecting retinal abnormalities in a real clinical setting. The images have been obtained by portable lens connected to a smart phone. In detecting the abnormalities related to diabetic retinopathy (DR), the algorithm provided 94.67 and 98.13% for true negatives and true positives, respectively.

Investigations of severity level measurements for diabetic macular oedema using machine learning algorithms

BACKGROUND

The macula is an important part of the human visual system and is responsible for clear and colour vision. Macular oedema happens when fluid and protein deposit on or below the macula of the eye and cause the macula to thicken and swell. Normally, it occurs due to diabetes called diabetic macular oedema. Diabetic macular oedema (DME) is one of the main causes of visual impairment in patients.

AIM

The aims of the present study are to detect and localize abnormalities in blood vessels with respect to macula in order to prevent vision loss for the diabetic patients.

METHODS

In this work, a novel fully computerized algorithm is used for the recognition of various diseases in macula using both fundus images and optical coherence tomography (OCT) images. Abnormal blood vessels are segmented using thresholding algorithm. The classification is performed by three different classifiers, namely, the support vector machine (SVM), cascade neural network (CNN) and partial least square (PLS) classifiers, which are employed to identify whether the image is normal or abnormal.

CONCLUSION

The results of all of the classifiers are compared based on their accuracy. The classifier accuracies of the SVM, cascade neural network and partial least square are 98.33, 97.16 and 94.34%, respectively. While analysing DME using both images, OCT produced efficient output than fundus images. Information about the severity of the disease and the localization of the pathologies is very useful to the ophthalmologist for diagnosing disease and choosing the proper treatment for a patient to prevent vision loss.

Localization of optic disc and fovea in retinal images using intensity based line scanning analysis

Accurate detection of diabetic retinopathy (DR) mainly depends on identification of retinal landmarks such as optic disc and fovea. Present methods suffer from challenges like less accuracy and high computational complexity. To address this issue, this paper presents a novel approach for fast and accurate localization of optic disc (OD) and fovea using one-dimensional scanned intensity profile analysis. The proposed method utilizes both time and frequency domain information effectively for localization of OD. The final OD center is located using signal peak-valley detection in time domain and discontinuity detection in frequency domain analysis. However, with the help of detected OD location, the fovea center is located using signal valley analysis. Experiments were conducted on MESSIDOR dataset, where OD was successfully located in 1197 out of 1200 images (99.75%) and fovea in 1196 out of 1200 images (99.66%) with an average computation time of 0.52s. The large scale evaluation has been carried out extensively on nine publicly available databases. The proposed method is highly efficient in terms of quickly and accurately localizing OD and fovea structure together compared with the other state-of-the-art methods.

A review on automatic analysis techniques for color fundus photographs

In this paper, we give a review on automatic image processing tools to recognize diseases causing specific distortions in the human retina. After a brief summary of the biology of the retina, we give an overview of the types of lesions that may appear as biomarkers of both eye and non-eye diseases. We present several state-of-the-art procedures to extract the anatomic components and lesions in color fundus photographs and decision support methods to help clinical diagnosis. We list publicly available databases and appropriate measurement techniques to compare quantitatively the performance of these approaches. Furthermore, we discuss on how the performance of image processing-based systems can be improved by fusing the output of individual detector algorithms. Retinal image analysis using mobile phones is also addressed as an expected future trend in this field.

Optic Disc Localization Using Directional Models

Reliable localization of the optic disc (OD) is important for retinal image analysis and ophthalmic pathology screening. This paper presents a novel method to automatically localize ODs in retinal fundus images based on directional models. According to the characteristics of retina vessel networks, such as their origin at the OD and parabolic shape of the main vessels, a global directional model, named the relaxed biparabola directional model, is first built. In this model, the main vessels are modeled by using two parabolas with a shared vertex and different parameters. Then, a local directional model, named the disc directional model, is built to characterize the local vessel convergence in the OD as well as the shape and the brightness of the OD. Finally, the global and the local directional models are integrated to form a hybrid directional model, which can exploit the advantages of the global and local models for highly accurate OD localization. The proposed method is evaluated on nine publicly available databases, and achieves an accuracy of 100% for each database, which demonstrates the effectiveness of the proposed OD localization method.

Retinal image quality assessment based on image clarity and content

Retinal image quality assessment (RIQA) is an essential step in automated screening systems to avoid misdiagnosis caused by processing poor quality retinal images. A no-reference transform-based RIQA algorithm is introduced that assesses images based on five clarity and content quality issues: sharpness, illumination, homogeneity, field definition, and content. Transform-based RIQA algorithms have the advantage of considering retinal structures while being computationally inexpensive. Wavelet-based features are proposed to evaluate the sharpness and overall illumination of the images. A retinal saturation channel is designed and used along with wavelet-based features for homogeneity assessment. The presented sharpness and illumination features are utilized to assure adequate field definition, whereas color information is used to exclude nonretinal images. Several publicly available datasets of varying quality grades are utilized to evaluate the feature sets resulting in area under the receiver operating characteristic curve above 0.99 for each of the individual feature sets. The overall quality is assessed by a classifier that uses the collective features as an input vector. The classification results show superior performance of the algorithm in comparison to other methods from literature. Moreover, the algorithm addresses efficiently and comprehensively various quality issues and is suitable for automatic screening systems.

A novel method for retinal exudate segmentation using signal separation algorithm

Diabetic retinopathy is one of the major causes of blindness in the world. Early diagnosis of this disease is vital to the prevention of visual loss. The analysis of retinal lesions such as exudates, microaneurysms and hemorrhages is a prerequisite to detect diabetic disorders such as diabetic retinopathy and macular edema in fundus images. This paper presents an automatic method for the detection of retinal exudates. The novelty of this method lies in the use of Morphological Component Analysis (MCA) algorithm to separate lesions from normal retinal structures to facilitate the detection process. In the first stage, vessels are separated from lesions using the MCA algorithm with appropriate dictionaries. Then, the lesion part of retinal image is prepared for the detection of exudate regions. The final exudate map is created using dynamic thresholding and mathematical morphologies. Performance of the proposed method is measured on the three publicly available DiaretDB, HEI-MED and e-ophtha datasets. Accordingly, the AUC of 0.961 and 0.948 and 0.937 is achieved respectively, which are greater than most of the state-of-the-art methods.

Mediastinal lymph node detection and station mapping on chest CT using spatial priors and random forest

PURPOSE

To develop an automated system for mediastinal lymph node detection and station mapping for chest CT.

METHODS

The contextual organs, trachea, lungs, and spine are first automatically identified to locate the region of interest (ROI) (mediastinum). The authors employ shape features derived from Hessian analysis, local object scale, and circular transformation that are computed per voxel in the ROI. Eight more anatomical structures are simultaneously segmented by multiatlas label fusion. Spatial priors are defined as the relative multidimensional distance vectors corresponding to each structure. Intensity, shape, and spatial prior features are integrated and parsed by a random forest classifier for lymph node detection. The detected candidates are then segmented by the following curve evolution process. Texture features are computed on the segmented lymph nodes and a support vector machine committee is used for final classification. For lymph node station labeling, based on the segmentation results of the above anatomical structures, the textual definitions of mediastinal lymph node map according to the International Association for the Study of Lung Cancer are converted into patient-specific color-coded CT image, where the lymph node station can be automatically assigned for each detected node.

RESULTS

The chest CT volumes from 70 patients with 316 enlarged mediastinal lymph nodes are used for validation. For lymph node detection, their system achieves 88% sensitivity at eight false positives per patient. For lymph node station labeling, 84.5% of lymph nodes are correctly assigned to their stations.

CONCLUSIONS

Multiple-channel shape, intensity, and spatial prior features aggregated by a random forest classifier improve mediastinal lymph node detection on chest CT. Using the location information of segmented anatomic structures from the multiatlas formulation enables accurate identification of lymph node stations.

Localization and segmentation of optic disc in retinal images using circular Hough transform and grow-cut algorithm

Automated retinal image analysis has been emerging as an important diagnostic tool for early detection of eye-related diseases such as glaucoma and diabetic retinopathy. In this paper, we have presented a robust methodology for optic disc detection and boundary segmentation, which can be seen as the preliminary step in the development of a computer-assisted diagnostic system for glaucoma in retinal images. The proposed method is based on morphological operations, the circular Hough transform and the grow-cut algorithm. The morphological operators are used to enhance the optic disc and remove the retinal vasculature and other pathologies. The optic disc center is approximated using the circular Hough transform, and the grow-cut algorithm is employed to precisely segment the optic disc boundary. The method is quantitatively evaluated on five publicly available retinal image databases DRIVE, DIARETDB1, CHASE_DB1, DRIONS-DB, Messidor and one local Shifa Hospital Database. The method achieves an optic disc detection success rate of 100% for these databases with the exception of 99.09% and 99.25% for the DRIONS-DB, Messidor, and ONHSD databases, respectively. The optic disc boundary detection achieved an average spatial overlap of 78.6%, 85.12%, 83.23%, 85.1%, 87.93%, 80.1%, and 86.1%, respectively, for these databases. This unique method has shown significant improvement over existing methods in terms of detection and boundary extraction of the optic disc.

Optic Disc and Optic Cup Segmentation Methodologies for Glaucoma Image Detection: A Survey

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 segmentation techniques are used to isolate the relevant parts of the retinal image and to calculate the cup-to-disc ratio. The main objective of this paper is to review segmentation methodologies and techniques for the disc and cup boundaries which are utilized to calculate the disc and cup geometrical parameters automatically and accurately to help the professionals in the glaucoma to have a wide view and more details about the optic nerve head structure using retinal fundus images. We provide a brief description of each technique, highlighting its classification and performance metrics. The current and future research directions are summarized and discussed.