Retinal Image Quality Assessment Research Articles

The Assessment of Retinal Image Quality Using a Non-Mydriatic Fundus Camera in a Teleophthalmologic Platform.

This study assesses the quality of retinal images captured using a non-mydriatic fundus camera within a teleophthalmologic platform in Taiwan. The objective was to evaluate the effectiveness of non-mydriatic fundus cameras for remote retinal screening and identify factors impacting image quality. From June 2020 to August 2022, 629 patients from five rural infirmaries underwent ophthalmic examinations, with fundus images captured without pupil dilation. These images were reviewed by senior ophthalmologists and graded based on quality. The results indicated that approximately 70% of images were of satisfactory diagnostic quality. Risk factors for poor image quality included older age, the presence of cataracts, pseudophakia, and diabetes mellitus. This study demonstrates the feasibility of using non-mydriatic fundus cameras for teleophthalmology, highlighting the importance of identifying and addressing factors that affect image quality to enhance diagnostic accuracy in remote settings.

No reference retinal image quality assessment using support vector machine classifier in wavelet domain

No reference retinal image quality assessment using support vector machine classifier in wavelet domain

Enhancing Retinal Fundus Image Quality Assessment With Swin-Transformer-Based Learning Across Multiple Color-Spaces.

The assessment of retinal image (RI) quality holds significant importance in both clinical trials and large datasets, because suboptimal images can potentially conceal early signs of diseases, thereby resulting in inaccurate medical diagnoses. This study aims to develop an automatic method for Retinal Image Quality Assessment (RIQA) that incorporates visual explanations, aiming to comprehensively evaluate the quality of retinal fundus images (RIs). We developed an automatic RIQA system, named Swin-MCSFNet, utilizing 28,792 RIs from the EyeQ dataset, as well as 2000 images from the EyePACS dataset and an additional 1,000 images from the OIA-ODIR dataset. After preprocessing, including cropping black regions, data augmentation, and normalization, a Swin-MCSFNet classifier based on the Swin-Transformer for multiple color-space fusion was proposed to grade the quality of RIs. The generalizability of Swin-MCSFNet was validated across multiple data centers. Additionally, for enhanced interpretability, a Score-CAM-generated heatmap was applied to provide visual explanations. Experimental results reveal that the proposed Swin-MCSFNet achieves promising performance, yielding a micro-receiver operating characteristic (ROC) of 0.93 and ROC scores of 0.96, 0.81, and 0.96 for the "Good," "Usable," and "Reject" categories, respectively. These scores underscore the accuracy of RIQA based on Swin-MCSF in distinguishing among the three categories. Furthermore, heatmaps generated across different RIQA classification scores and various color spaces suggest that regions in the retinal images from multiple color spaces contribute significantly to the decision-making process of the Swin-MCSFNet classifier. Our study demonstrates that the proposed Swin-MCSFNet outperforms other methods in experiments conducted on multiple datasets, as evidenced by the superior performance metrics and insightful Score-CAM heatmaps. This study constructs a new retinal image quality evaluation system, which will contribute to the subsequent research of retinal images.

Robust deep learning for eye fundus images: Bridging real and synthetic data for enhancing generalization

Deep learning applications for assessing medical images are limited because the datasets are often small and imbalanced. The use of synthetic data has been proposed in the literature, but neither a robust comparison of the different methods nor generalizability has been reported. Our approach integrates a retinal image quality assessment model and StyleGAN2 architecture to enhance Age-related Macular Degeneration (AMD) detection capabilities and improve generalizability. This work compares ten different Generative Adversarial Network (GAN) architectures to generate synthetic eye-fundus images with and without AMD. We combined subsets of three public databases (iChallenge-AMD, ODIR-2019, and RIADD) to form a single training and test set. We employed the STARE dataset for external validation, ensuring a comprehensive assessment of the proposed approach. The results show that StyleGAN2 reached the lowest Fréchet Inception Distance (166.17), and clinicians could not accurately differentiate between real and synthetic images. ResNet-18 architecture obtained the best performance with 85% accuracy and outperformed the two human experts (80% and 75%) in detecting AMD fundus images. The accuracy rates were 82.8% for the test set and 81.3% for the STARE dataset, demonstrating the model’s generalizability. The proposed methodology for synthetic medical image generation has been validated for robustness and accuracy, with free access to its code for further research and development in this field.

Impact of lens autofluorescence and opacification on retinal imaging

BackgroundRetinal imaging, including fundus autofluorescence (FAF), strongly depends on the clearness of the optical media. Lens status is crucial since the ageing lens has both light-blocking and autofluorescence (AF) properties...

A deep retinal image quality assessment network with salient structure priors

Retinal image quality assessment is an essential prerequisite for diagnosis of retinal diseases. Its goal is to identify high quality retinal images in which anatomic structures and lesions attracting ophthalmologists’attention most are exhibited clearly and definitely while reject poor quality images. Motivated by this, we mimic the way that ophthalmologists assess the quality of retinal images and propose a method termed SalStructuIQA. First, two salient structures are detected for retinal quality assessment. One is large-size salient structures including optic disc and exudates in large-size. The other is tiny-size salient structures mainly including vessels. Then the proposed two salient structure priors are incorporated with deep convolutional neural network (CNN) to enforce CNN pay attention to these salient structures. Accordingly, two CNN architectures named Dual-branch SalStructIQA and Single-branch SalStructIQA are designed for the incorporation, respectively. Experimental results show that the Fscore of our proposed Dual-branch SalStructIQA and Single-branch SalStructIQA are 0.8723 and 0.8662 respectively on the public Eye-Quality dataset, which demonstrates the effectiveness of our methods.

Learning for retinal image quality assessment with label regularization

Background and Objective: The assessment of the image quality is crucial before the computer-aided diagnosis of fundus images. This task is very challenging. Firstly, the subjective judgments of graders on image quality lead to ambiguous labels. Secondly, despite being treated as classification in existing works, grading has regression properties that cannot be ignored. Solving the ambiguity problem and regression problem in the label space, and extracting discriminative features, have become the keys to quality assessment.Methods: In this paper, we proposed a framework that can assess the quality of fundus images accurately and reasonably based on deep convolutional neural networks. Drawing on the experience of human graders, a dual-path convolutional neural network with attention blocks is designed to better extract discriminative features and present the bases of decision. Label smoothing and cost-sensitive regularization are designed to solve the label ambiguity problem and the potential regression problem respectively. Besides, a large number of images are annotated by us to further improve the results.Results: We conducted our experiments on the largest retinal image quality assessment dataset with 28,792 retinal images. Our approach achieves 0.8868 precision, 0.8786 recall, 0.8820 F1, and 0.9138 Kappa score. Results show that our approach outperforms state-of-the-art methods.Conclusions: The promising performances reveal that our methods are beneficial to retinal image quality assessment and have potential in other grading tasks.

Suitability Classification of Retinal Fundus Images for Diabetic Retinopathy Using Deep Learning

Diabetic retinopathy (DR) is the primary cause of blindness in developing and developed countries. Early-stage DR detection reduces the risk of blindness in Diabetes Mellitus (DM) patients. There has been a sharp rise in the prevalence of DM in recent years, especially in low- and middle-income countries. In this context, automated artificial intelligence-based DM screening is a crucial tool to help classify the considerable amount of Retinal Fundus Images (RFI). However, retinal image quality assessment has shown to be fundamental in real-world DR screening processes to avoid out-of-distribution data, drift, and images lacking relevant anatomical information. This work analyzes the spatial domain features and image quality assessment metrics for carrying out Deep Learning (DL) classification and detecting notable features in RFI. In addition, a novel lightweight convolutional neural network is proposed specifically for binary classification at a low computational cost. The training results are comparable to state-of-the-art neural networks, which are widely used in DL applications. The implemented architecture achieves 98.6% area under the curve, and 97.66%, and 98.33% sensitivity and specificity, respectively. Moreover, the object detection model trained achieves 94.5% mean average precision. Furthermore, the proposed approach can be integrated into any automated RFI analysis system.

A dark and bright channel prior guided deep network for retinal image quality assessment

Background and objectiveRetinal image quality assessment is an essential task for the diagnosis of retinal diseases. Recently, there are emerging deep models to grade quality of retinal images. However, current models either directly transfer classification networks originally designed for natural images to quality classification of retinal images or introduce extra image quality priors via multiple CNN branches or independent CNNs. The purpose of this work is to address retinal image quality assessment by a simple deep model. MethodsWe propose a dark and bright channel prior guided deep network for retinal image quality assessment named GuidedNet. It introduces dark and bright channel priors into deep network without extra parameters increasing and allows for training end-to-end. In detail, the dark and bright channel priors are embedded into the start layer of a deep network to improve the discriminate ability of deep features. Moreover, we re-annotate a new retinal image quality dataset called RIQA-RFMiD for further validation. ResultsThe proposed method is evaluated on a public retinal image quality dataset Eye-Quality and our re-annotated dataset RIQA-RFMiD. We obtain the average F-score of 88.03% on Eye-Quality and 66.13% on RIQA-RFMiD, respectively. ConclusionsWe investigate the utility of the dark and bright channel priors for retinal image quality assessment. And we propose a GuidedNet by embedding the dark and bright channel priors into CNNs without much model burden. Moreover, to valid the GuidedNet, we re-create a new dataset RIQA-RFMiD. With the GuidedNet, we achieves state-of-the-art performances on a public dataset Eye-Quality and our re-annotated dataset RIQA-RFMiD.

No-Reference Retinal Image Sharpness Metric Using Daubechies Wavelet Transform

Retinal fundus images are increasingly used by ophthalmologists both manually and without human intervention for detecting ocular diseases. Poor quality retinal images could lead to misdiagnosis or delayed treatment. Hence, a picture quality index was a crucial measure to ensure that the obtained images from acquisition system were suitable for reliable medical diagnosis. In this paper, a no-reference retinal image quality assessment based on wavelet transform is presented. A multiresolution Daubechies (db2) wavelet at level 4 was employed to decompose an original image into detail, and approximation sub-bands for extracting the sharpness information. The sharpness quality index was calculated from the entropy of the sub-bands. The proposed measure was validated by using images from the High-Resolution Fundus (HRF) dataset. The experimental results show that the proposed index performed more consistent with human visual perception and outperformed the Abdel-Hamid et al method.

Blood vessel and background separation for retinal image quality assessment

Blood vessel and background separation for retinal image quality assessment

Deep Learning for Retinal Image Quality Assessment of Optic Nerve Head Disorders.

Deep learning (DL)-based retinal image quality assessment (RIQA) algorithms have been gaining popularity, as a solution to reduce the frequency of diagnostically unusable images. Most existing RIQA tools target retinal conditions, with a dearth of studies looking into RIQA models for optic nerve head (ONH) disorders. The recent success of DL systems in detecting ONH abnormalities on color fundus images prompts the development of tailored RIQA algorithms for these specific conditions. In this review, we discuss recent progress in DL-based RIQA models in general and the need for RIQA models tailored for ONH disorders. Finally, we propose suggestions for such models in the future.

Retinal image quality assessment using transfer learning: Spatial images vs. wavelet detail subbands

Retinal image quality assessment (RIQA) is essential to assure that images used for medical analysis are of sufficient quality for reliable diagnosis. A modified VGG16 network with transfer learning is introduced in order to classify retinal images into good or bad quality images. Both spatial and wavelet detail subbands are compared as inputs to the modified VGG16 network. Three public retinal image datasets captured with different imaging devices are used, both individually and collectively. Superior performance was attained by the modified VGG16 network, where accuracies in the range of 99–100% were achieved regardless of whether retinal images from the same or different sources were considered and whether the spatial or wavelet images were used. The implemented RIQA algorithm was also found to outperform other RIQA deep learning algorithms from literature by 1.5–10% and to achieve accuracies that are up to 32% higher than traditional RIQA methods for the same dataset.

A Teleophthalmology Support System Based on the Visibility of Retinal Elements Using the CNNs.

This paper proposes a teleophthalmology support system in which we use algorithms of object detection and semantic segmentation, such as faster region-based CNN (FR-CNN) and SegNet, based on several CNN architectures such as: Vgg16, MobileNet, AlexNet, etc. These are used to segment and analyze the principal anatomical elements, such as optic disc (OD), region of interest (ROI) composed by the macular region, real retinal region, and vessels. Unlike the conventional retinal image quality assessment system, the proposed system provides some possible reasons about the low-quality image to support the operator of an ophthalmoscope and patient to acquire and transmit a better-quality image to central eye hospital for its diagnosis. The proposed system consists of four steps: OD detection, OD quality analysis, obstruction detection of the region of interest (ROI), and vessel segmentation. For the OD detection, artefacts and vessel segmentation, the FR-CNN and SegNet are used, while for the OD quality analysis, we use transfer learning. The proposed system provides accuracies of 0.93 for the OD detection, 0.86 for OD image quality, 1.0 for artefact detection, and 0.98 for vessel segmentation. As the global performance metric, the kappa-based agreement score between ophthalmologist and the proposed system is calculated, which is higher than the score between ophthalmologist and general practitioner.

Fast macula detection and application to retinal image quality assessment

In this article, we present a segmentation algorithm for assessing retinal image quality with respect to the visibility of the macular region. An image is considered of acceptable quality if the macular region is clearly visible and entirely in the field of view. Additionally, for acceptable images, the method is able to locate the fovea with a maximal error of 0.34 mm. The algorithm is based on a lightweight fully-convolutional network, several thousand times smaller than state-of-the-art networks investigated so far in preliminary studies. We obtain near-human performance for assessing macula visibility and fovea localization. The presented method can easily be embedded in tabletop or handheld retinographs, decreasing the number of ungradable images, saving both patient and physician time. It is an important step towards automatic screening of retinal pathologies, including diabetic retinopathy, which is a major global healthcare issue.

Retinal Image Quality Assessment: Portable Eye Examination Kit Retina (Peek Retina)Tm Versus 3d-Printed Ophthalmoscope (3dpo)

Portable Eye Examination Kit retina (Peek Retina™, Peek Vision Ltd, UK) and 3D Printed Ophthalmoscope (3DPO) were identified to have acceptable image for retinal evaluation, however the retinal images quality in term of blood vessels visibility between both devices was uncertain. This study was conducted to compare the quality of image based on blood vessels visibility between Peek Retina and 3DPO for fractal dimension (Df) analysis. In this study, a total of 40 retinal images (nPEEK=20, n3DPO=20) of 20 participants were captured on random eyes. The best retinal images with good focus and significant retinal blood vessels visibility of Peek Retina and 3DPO were selected for image quality analysis. The retinal images were cropped approximately following the size of the cornea and resized to 900 by 900 pixels of resolution using GNU Image Manipulation Program Version 2.8.18 (GIMP). The images were randomly sorted as Retinal Image Quality Assessment (RIQA) generated by Google form. Likert scale was implemented to assess the preferences scale of retinal image quality in determining the visibility of retinal vasculature to be traced with four choices of response options (1; very difficult, 2; difficult, 3; easy and 4; very easy). Prior to the retinal image assessment, ten optometrists were asked to experience retinal blood vessels tracing and consequently evaluate the 40 images by choosing the scale options (1 to 4) based on visibility retinal blood vessels. Mann-Whitney test indicated that the blood vessel tracing was easier for Peek Retina (median = 3) than for 3DPO (median = 2), p < 0.0001. Retinal image captured by Peek Retina was rated as very easy (43.5%) for blood vessels tracing as compared to the image from 3DPO (17.0%)Error! Reference source not found.. Only 1.5% of the image captured by PEEK was considered as a very difficult for blood vessel tracing. Difficult and easy preference scales of blood vessel tracing for PEEK were 16.5% and 38.5% respectively. 34% of 3DPO retinal image was graded as difficult for blood vessel tracing followed by 28.5% (easy), 20.5% (very difficult) and 17.0% (very easy). The results indicate that a retinal image photographed by Peek Retina was more preferable in tracing retinal vascular network for Df analysis as compared to 3DPO.

Retinal image quality assessment based on FA-Net

Objective To propose a deep learning-based retinal image quality classification network, FA-Net, to make convolutional neural network (CNN) more suitable for image quality assessment in eye disease screening system. Methods The main network of FA-Net was composed of VGG-19.On this basis, attention mechanism was added to the CNN.By using transfer learning method in training, the weight of ImageNet was used to initialize the network.The attention net is based on foreground extraction by extracting the blood vessel and suspected regions of lesion and assigning higher weights to region of interest to enhance the learning of these important areas. Results Total of 2 894 fundus images were used for training FA-Net.FA-Net achieved 97.65% classification accuracy on a test set containing 2 170 fundus images, with the sensitivity and specificity of 0.978 and 0.960, respectively, and the area under curve(AUC) was 0.995. Conclusions Compared with other CNNs, the proposed FA-Net has better classification performance and can evaluate retinal fundus image quality more accurately and efficiently.The network takes into account the human visual system (HVS) and human attention mechanism.By adding attention module into the VGG-19 network structure, the classification results can be better interpreted as well as better classification performance. Key words: Retinal image quality assessment; Retinal image enhancement; Convolutional neural networks; Deep learning

A CNN-BASED RETINAL IMAGE QUALITY ASSESSMENT SYSTEM FOR TELEOPHTHALMOLOGY

Oculopathy is a widespread disease among people of all ages around the world. Teleophthalmology can facilitate the ophthalmological diagnosis for less developed countries that lack medical resources. In teleophthalmology, the assessment of retinal image quality is of great importance. In this paper, we propose a no-reference retinal image assessment system based on DenseNet, a convolutional neural network architecture. This system classified fundus images into good and bad quality or five categories: adequate, just noticeable blur, inappropriate illumination, incomplete optic disc, and opacity. The proposed system was evaluated on different datasets and compared to the applications based on other two networks: VGG-16 and GoogLenet. For binary classification, the good-and-bad binary classifier achieves an AUC of 1.000, and the degradation-specified classifiers that distinguish one specified degradation versus the rest achieve AUC values of 0.972, 0.990, 0.982, 0.982 for four categories, respectively. The multi-classification based on DenseNet achieves an overall accuracy of 0.927, which is significantly higher than 0.549 and 0.757 obtained using VGG-16 and GoogLeNet, respectively. The experimental results indicate that the proposed approach produces outstanding performance in retinal image quality assessment and is worth applying in ophthalmological telemedicine applications. In addition, the proposed approach is robust to the image noise. This study fills the gap of multi-classification in retinal image quality assessment.

Retinal image quality assessment for diabetic retinopathy screening: A survey

Retinal image quality assessment (RIQA) is one of the key components in screening for diabetic retinopathy (DR). As one of the most serious complications of diabetes, DR has become a leading cause of blindness in adults globally. DR screening is essential to achieve early diagnosis so that effective treatment could be provided timely. However, the collected images of medically unsatisfactory quality always lead to failure of diagnosis and waste of ophthalmologists’ precious time. Hence, the first step in a good DR screening program is verifying retinal images of good quality. In this paper, we provide a systematic review on automated assessment of retinal image quality for DR screening. Scheme and parameters for RIQA are firstly presented. Next, we provide detailed understanding of the existing RIQA techniques, algorithms and methodologies, including brief description and analysis of each existing state-of-art approaches and comparison between such methods. Datasets and evaluation metrics are also illustrated. Finally, several challenges and future research directions are summarized and discussed.

Combination of Global Features for the Automatic Quality Assessment of Retinal Images.

Diabetic retinopathy (DR) is one of the most common causes of visual loss in developed countries. Computer-aided diagnosis systems aimed at detecting DR can reduce the workload of ophthalmologists in screening programs. Nevertheless, a large number of retinal images cannot be analyzed by physicians and automatic methods due to poor quality. Automatic retinal image quality assessment (RIQA) is needed before image analysis. The purpose of this study was to combine novel generic quality features to develop a RIQA method. Several features were calculated from retinal images to achieve this goal. Features derived from the spatial and spectral entropy-based quality (SSEQ) and the natural images quality evaluator (NIQE) methods were extracted. They were combined with novel sharpness and luminosity measures based on the continuous wavelet transform (CWT) and the hue saturation value (HSV) color model, respectively. A subset of non-redundant features was selected using the fast correlation-based filter (FCBF) method. Subsequently, a multilayer perceptron (MLP) neural network was used to obtain the quality of images from the selected features. Classification results achieved 91.46% accuracy, 92.04% sensitivity, and 87.92% specificity. Results suggest that the proposed RIQA method could be applied in a more general computer-aided diagnosis system aimed at detecting a variety of retinal pathologies such as DR and age-related macular degeneration.