Color Fundus Images Research Articles

Regional choroidal thickness estimation from color fundus images based on convolutional neural networks

PurposeThis study aims to estimate the regional choroidal thickness from color fundus images from convolutional neural networks in different network structures and task learning models. Method1276 color fundus photos and their corresponding choroidal thickness values from healthy subjects were obtained from the Topcon DRI Triton optical coherence tomography machine. Initially, ten commonly used convolutional neural networks were deployed to identify the most accurate model, which was subsequently selected for further training. This selected model was then employed in combination with single-, multiple-, and auxiliary-task training models to predict the average and sub-region choroidal thickness in both ETDRS (Early Treatment Diabetic Retinopathy Study) grids and 100-grid subregions. The values of mean absolute error and coefficient of determination (R2) were involved to evaluate the models' performance. ResultsEfficientnet-b0 network outperformed other networks with the lowest mean absolute error value (25.61 μm) and highest R2 (0.7817) in average choroidal thickness. Incorporating diopter spherical, anterior chamber depth, and lens thickness as auxiliary tasks improved predicted accuracy (p-value = 6.39×10−44, 2.72×10−38, 1.15×10−36 respectively). For ETDRS regional choroidal thickness estimation, multi-task model achieved better results than single task model (lowest mean absolute error = 31.10 μm vs. 33.20 μm). The multi-task training also can simultaneously predict the choroidal thickness of 100 grids with a minimum mean absolute error of 33.86 μm. ConclusionsEfficientnet-b0, in combination with multi-task and auxiliary task models, achieve high accuracy in estimating average and regional macular choroidal thickness directly from color fundus photographs.

FA4SANS-GAN: A Novel Machine Learning Generative Adversarial Network to Further Understand Ophthalmic Changes in Spaceflight Associated Neuro-Ocular Syndrome (SANS)

PurposeTo provide automated system for synthesizing fluorescein angiography (FA) images from color fundus (CF) photographs for averting risks associated with fluorescein dye and extend its future application to Space-associated neuro-ocular syndrome (SANS) detection in spaceflight where resources are limited. DesignDevelopment and validation of a novel conditional generative adversarial network (GAN) trained on limited amount of fluorescein angiography and color fundus images with diabetic retinopathy and control cases. ParticipantsColor fundus and fluorescein angiography paired images for unique patients were collected from a publicly available study. MethodsFA4SANS-GAN was trained to generate fluorescein angiography images from color fundus photographs using two multi-scale generators coupled with two patch-GAN discriminators. 850 CF and FA images were utilized for training by augmenting images from 17 unique patients. The model was evaluated on 56 fluorescein images collected from 14 unique patients. In addition, it was compared with three other GAN architectures trained on the same dataset. Furthermore, we test the robustness of the models against acquisition noise and retaining structural information when introduced to artificially created biological markers. Main Outcome MeasureFor GAN synthesis metric Fréchet Inception Distance (FID) and Kernel Inception Distance (KID). Also, two one-sided tests (TOST) based on Welch's t-test for measuring statistical significance. ResultsOn test FA images, mean FID for FA4SANS-GAN was 39.8 (SD 9.9), which is better than GANgio model’s mean of 43.2 (SD 13.7), Pix2PixHD’s mean of 57.3 (SD 11.5) and Pix2Pix’s mean of 67.5 (SD 11.7). Similarly for KID, FA4SANS-GAN achieved mean of 0.00278 (SD 0.00167) which is better than other three model’s mean KID of 0.00303 (SD 0.00216), 0.00609 (SD 0.00238), 0.00784 (SD 0.00218). For TOST measurement, FA4SANS-GAN has proven to be statistically significant vs. GANgio (p = 0.006); vs Pix2PixHD (p < 0.00001); and vs. Pix2Pix (p < 0.00001). ConclusionOur study has shown FA4SANS-GAN to be statistically significant for two GAN synthesis metrics. Moreover, it is robust against acquisition noise, and can retain clear biological markers compared to the other three GAN architectures. This deployment of this model can be crucial in International Space Station (ISS) for detecting Spaceflight-associated neuro-ocular syndrome (SANS).

Advancements in Glaucoma Diagnosis: A Comprehensive Literature Survey on CNN-based Deep Learning Approaches

This literature survey review explores advancements in glaucoma diagnosis using convolutional neural networks (CNNs) within the realm of deep learning (DL). Glaucoma, a chronic and irreversible eye disease leading to vision deterioration, poses a significant global health challenge. Traditional diagnosis through colour fundus images is time-consuming, requiring skilled clinicians. The paper examines the development of a six-layered CNN architecture, integrating dropout and data augmentation techniques to enhance diagnostic accuracy. Focused on identifying intricate features like microaneurysms, exudate, and hemorrhages on the retina, this CNN-based approach offers a streamlined and efficient alternative to manual diagnosis. With glaucoma being a leading cause of blindness worldwide, the proposed methodology, trained on high-performance GPUs, presents a promising avenue for improving diagnostic efficiency and accuracy, thereby contributing to the evolution of glaucoma diagnosis

Analysis of the Retinal and Choroidal Vasculature Using Ultrawidefield Fundus Imaging in Mild Cognitive Impairment and Normal Cognition

PurposeTo utilize ultra-widefield (UWF) imaging to evaluate retinal and choroidal vasculature and structure in individuals with mild cognitive impairment (MCI) compared to controls with normal cognition. DesignProspective cross-sectional study ParticipantsOne hundred thirty-one eyes of 82 MCI patients and 231 eyes of 133 cognitively normal participants from the Eye Multimodal Imaging in Neurodegenerative Disease (iMIND) Study. MethodsA scanning laser ophthalmoscope (SLO) (California, Optos Inc, Marlborough, MA) was used to obtain UWF fundus color images. Images were analyzed with the Vasculature Assessment Platform for Images of the Retina UWF (VAMPIRE-UWF 2.0, Universities of Edinburgh and Dundee, United Kingdom) software. Main outcome measuresImaging parameters included vessel width gradient, vessel width intercept, large vessel choroidal vascular density, vessel tortuosity, and vessel fractal dimension. ResultsBoth retinal artery and vein width gradients were less negative in MCI patients compared to controls, demonstrating decreased rates of vessel thinning at the periphery (p < 0.001, p = 0.027). Retinal artery and vein width intercepts, a metric that extrapolates the width of vessel at the center of the optic disc, were smaller in MCI patients compared to controls (p < 0.001, p = 0.017). The large vessel choroidal vascular density, which quantifies the vascular area versus total choroidal area, was greater in MCI patients compared to controls (p = 0.025). ConclusionsWhen compared to controls with normal cognition, MCI patients had thinner retinal vasculature manifested in both the retinal arteries and in the veins. In MCI, these thinner arteries and veins attenuated at a lower rate when traveling towards the periphery. MCI patients also had increased choroidal vascular density.

On Performance Analysis Of Diabetic Retinopathy Classification

This paper describes the Classification of bulk OCT retinal fundus images of normal and diabetic retinopathy using the Intensity histogram features, Gray Level Co-Occurrence Matrix (GLCM), and the Gray Level Run Length Matrix (GLRLM) feature extraction techniques. Three features—Intensity histogram features, GLCM, and GLRLM were taken and, that features were compared fairly. A total of 301 bulk OCT retinal fundus color images were taken for two different varieties which are normal and diabetic retinopathy. For classification and feature extraction, a filtered image output based on a fourth-order PDE is used. Using OCT retinal fundus images, the most effective feature extraction method is identified.

A deep learning framework for the early detection of multi-retinal diseases.

Retinal images play a pivotal contribution to the diagnosis of various ocular conditions by ophthalmologists. Extensive research was conducted to enable early detection and timely treatment using deep learning algorithms for retinal fundus images. Quick diagnosis and treatment planning can be facilitated by deep learning models' ability to process images rapidly and deliver outcomes instantly. Our research aims to provide a non-invasive method for early detection and timely eye disease treatment using a Convolutional Neural Network (CNN). We used a dataset Retinal Fundus Multi-disease Image Dataset (RFMiD), which contains various categories of fundus images representing different eye diseases, including Media Haze (MH), Optic Disc Cupping (ODC), Diabetic Retinopathy (DR), and healthy images (WNL). Several pre-processing techniques were applied to improve the model's performance, such as data augmentation, cropping, resizing, dataset splitting, converting images to arrays, and one-hot encoding. CNNs have extracted extract pertinent features from the input color fundus images. These extracted features are employed to make predictive diagnostic decisions. In this article three CNN models were used to perform experiments. The model's performance is assessed utilizing statistical metrics such as accuracy, F1 score, recall, and precision. Based on the results, the developed framework demonstrates promising performance with accuracy rates of up to 89.81% for validation and 88.72% for testing using 12-layer CNN after Data Augmentation. The accuracy rate obtained from 20-layer CNN is 90.34% for validation and 89.59% for testing with Augmented data. The accuracy obtained from 20-layer CNN is greater but this model shows overfitting. These accuracy rates suggested that the deep learning model has learned to distinguish between different eye disease categories and healthy images effectively. This study's contribution lies in providing a reliable and efficient diagnostic system for the simultaneous detection of multiple eye diseases through the analysis of color fundus images.

External testing of an established glaucoma screening AI on a glaucoma clinic population from Finland

Purpose: To assess the external performance of an existing deep learning regression model for glaucoma detection (G‐RISK) in a glaucoma clinic population from Finland.Methods: Colour fundus images were retrospectively collected from the Tampere University Hospital (Tays), Finland. Eyes that had a visit in 2019 or 2020 were included based on the availability of ground truth labels. Ground truth labels for glaucoma were obtained through clinical evaluation of the optic nerve head (ONH) from colour fundus photos, clinical evaluation of retinal nerve fibre layer (RNFL) defects from fundus photos imaged with a cobalt filter, and clinical evaluation of the SITA Fast visual field (VF) test. Glaucoma diagnosis at Tays followed the ‘2 out of 3 rule’ recommended by the Finnish Evidence‐Based Guideline for Glaucoma, which requires at least 2 positive indications of glaucoma tests for referral [1]. Preprocessed colour fundus images were used as input to a previously trained deep learning model (G‐RISK) [2]. Performance was evaluated using the area under the receiver operating characteristic curve (AUC), accompanied by 95% confidence intervals. Patient aggregated results were obtained by comparing the maximum of both prediction and ground truth label.Results: Over 3000 patients had at least one ground truth available for a visit in 2019 or 2020. Clinical evaluation of structural features (ONH, RNFL) triggered more glaucomatous symptoms than the SITA Fast VF test. When thresholding the AI prediction against a single ground truth type, the best AUC was obtained on the ONH label (0.90 [0.89–0.91]), followed by RNFL (0.84 [0.83–0.86]) and VF (0.83 [0.81–0.84]). When implementing the 2 out of 3 rule, AUC reached 0.88 [0.82–0.90].Conclusion: The model performed best when using the ONH label and incorporating the ‘2 out of 3 rule’, reaching an AUC of 0.88. These findings suggest that the G‐RISK model could be a valuable tool for screening in clinical practice.

Clinically significant macular edema in an underserved population: Association with demographic factors and hemoglobin A1c.

Suspected clinically significant macular edema (SCSME) from exudates differed among ethnic groups in our underserved population. African American and Asian subjects had higher prevalence than Hispanics and non-Hispanic Caucasians, from the same clinics. Men had higher prevalence than women. Highly elevated blood glucose was frequent and associated with SCSME. We investigated the association between the presence of SCSME from exudates and hemoglobin A1c (HbA1c), as well as demographic factors such as age, sex, and ethnic group. Our population was underserved diabetic patients from the same geographic locations. Ethnic groups were White Hispanic, non-Hispanic Caucasian, African American, and Asian, with a high proportion of underrepresented minorities. In a diabetic retinopathy screening study at four community clinics in Alameda County, California, nonmydriatic 45° color fundus images were collected from underserved diabetic subjects following the EyePACS imaging protocol. Images were analyzed for SCSME from exudates by two certified graders. Logistic regression assessed the association between SCSME from exudates and age, sex, ethnic group, and HbA1c. Of 1997 subjects, 147 (7.36%) had SCSME from exudates. The mean ± standard deviation age was 53.4 ± 10.5 years. The mean ± standard deviation HbA1c level was 8.26 ± 2.04. Logistic regression analysis indicated a significant association between presence of SCSME from exudates and HbA1c levels (p<0.001), sex (p=0.027), and ethnicity (p=0.030). African Americans (odds ratio [OR], 1.63; 95% confidence interval [CI], 1.06 to 2.50; p=0.025) and Asians (OR, 1.63; 95% CI, 1.05 to 2.54; p=0.029) had a higher risk than Hispanics. After adjusting for ethnicity, sex, and age, the odds of developing SCSME from exudates increased by 26.5% with every 1% increase in HbA1c level (OR, 1.26; 95% CI, 1.18 to 1.36; p<0.001). In our underserved population, many diabetic patients had very high HbA1c values. Ethnic background (African American > Asians > Hispanics), sex (male > female), and HbA1c level were strong indicators for identifying who is at increased risk of developing SCSME from exudates.

Machine Learning Techniques used for Recognizing the Diabetic Retinopathy: A Survey

Diabetic Retinopathy (DR) is one of the major causes of blindness. DR mutilates the retinal blood vessels of a patient having diabetes. The DR has two major types: First one is Non- Proliferative Diabetic Retinopathy (NPDR) and second is Proliferative Diabetic Retinopathy (PDR). PDR is the advanced stage of DR which leads to neo vascularization, it is expected that the number of DR patients is to increase from 382 million to 592 million by 2028. In the early stages of the DR the patients were asymptomatic but in advanced stages, it leads to floaters, blurred vision, distortions, and progressive visual acuity loss. It is difficult but utmost important to detect the DR in early stages to avoid the worse effect of latter stages. The colour fundus images were used for the diagnosis of DR, the manual analysis could only be done by highly trained domain experts but it is bit expensive in terms of time and cost. Hence, it is important to use computer vision methods to automatically analyse the fundus images of Retina and assist the physicians/radiologists. The computer vision-based methods are divided into hand-on engineering and end-to-end learning. The hand-on engineering methods extract features using traditional approaches such as HoG, SIFT, LBP, Gabor filters, which failed to encode the variations in scale, rotation, and illumination. The end-to-end leaning automatically learns the hidden rich features and thus performs better classification.

Refined image quality assessment for color fundus photography based on deep learning.

Color fundus photography is widely used in clinical and screening settings for eye diseases. Poor image quality greatly affects the reliability of further evaluation and diagnosis. In this study, we developed an automated assessment module for color fundus photography image quality assessment using deep learning. A total of 55,931 color fundus photography images from multiple centers in Shanghai and the public database were collected and annotated as training, validation, and testing data sets. The pre-diagnosis image quality assessment module based on the multi-task deep neural network was designed. The detailed criterion of color fundus photography image quality including three subcategories with three levels of grading was applied to improve precision and objectivity. The auxiliary tasks such as the localization of the optic nerve head and macula, the classification of laterality, and the field of view were also included to assist the quality assessment. Finally, we validated our module internally and externally by evaluating the area under the receiver operating characteristic curve, sensitivity, specificity, accuracy, and quadratic weighted Kappa. The "Location" subcategory achieved area under the receiver operating characteristic curves of 0.991, 0.920, and 0.946 for the three grades, respectively. The "Clarity" subcategory achieved area under the receiver operating characteristic curves of 0.980, 0.917, and 0.954 for the three grades, respectively. The "Artifact" subcategory achieved area under the receiver operating characteristic curves of 0.976, 0.952, and 0.986 for the three grades, respectively. The accuracy and Kappa of overall quality reach 88.15% and 89.70%, respectively, on the internal set. These two indicators on the external set were 86.63% and 88.55%, respectively, which were very close to that of the internal set. This work showed that our deep module was able to evaluate the color fundus photography image quality using more detailed three subcategories with three grade criteria. The promising results on both internal and external validation indicated the strength and generalizability of our module.

Transforming Diabetic Retinopathy Detection: Advanced Strategies with Convolutional Neural Networks

Diabetic retinopathy is an aspect of diabetes which produces vascular abnormalities that might result in blindness. Since the symptoms with this illness are lasting, detection at an early stage is crucial because untreated eye disease may lead to blindness. The detection of microaneurysms using digital color fundus images is an essential initial step in automated diabetic retinopathy testing. Five different DR stages or classes (proliferative diabetic retinopathy), typical, mild, moderate, serious, and PDR, are typically assessed by expert doctors to determine this disastrous disorder. Manually diagnosis (by physicians) for such ailment requires time and is vulnerable to inaccuracy. There have been several computer vision-based algorithms created for autonomously recognizing DR and its different stages from retina scans. The results of this research utilize fundus-colored images to provide a computational method for recognizing and categorizing the condition via a convolution neural network.

Detection of Papilledema Severity from Color Fundus Images using Transfer Learning Approaches

Papilledema is edema in the area where the optic nerve meets the eye as a result of increased pressure inside the head. This disease can result in very serious problems, such as abnormal optical changes, decreased visual acuity, and even permanent blindness if left untreated. In this study, an image processing based solution was presented for the detection of papilledema severity from color fundus images using transfer learning approaches. The image dataset includes 295 papilledema images, 295 pseudopapilledema images, and 779 control images. Histogram equalization and the 3D box filter were used for image preprocessing. The images were enhanced with the histogram equalization method and denoised with the 3D box filter method. Then, the performances of EfficentNet-B0, GoogLeNet, MobileNetV2, NASNetMobile, and ResNet-101 transfer learning approaches were compared. The hold-out method was used to calculate the performance of transfer learning. In the experiments, the MobileNetV2 approach had the highest performance with 0.96 overall accuracy and 0.94 Cohen's Kappa. The results of the experiments proved that the combination of the histogram equalization, the 3D box filter, and the MobileNetV2 transfer learning approach can be used for automatic detection of papilledema severity. Compared to other similar studies that are known in the literature, the overall accuracy was higher.

Explainable artificial intelligence for the automated assessment of the retinal vascular tortuosity.

Retinal vascular tortuosity is an excessive bending and twisting of the blood vessels in the retina that is associated with numerous health conditions. We propose a novel methodology for the automated assessment of the retinal vascular tortuosity from color fundus images. Our methodology takes into consideration several anatomical factors to weigh the importance of each individual blood vessel. First, we use deep neural networks to produce a robust extraction of the different anatomical structures. Then, the weighting coefficients that are required for the integration of the different anatomical factors are adjusted using evolutionary computation. Finally, the proposed methodology also provides visual representations that explain the contribution of each individual blood vessel to the predicted tortuosity, hence allowing us to understand the decisions of the model. We validate our proposal in a dataset of color fundus images providing a consensus ground truth as well as the annotations of five clinical experts. Our proposal outperforms previous automated methods and offers a performance that is comparable to that of the clinical experts. Therefore, our methodology demonstrates to be a viable alternative for the assessment of the retinal vascular tortuosity. This could facilitate the use of this biomarker in clinical practice and medical research.

Explainable artificial intelligence model for the detection of geographic atrophy using colour retinal photographs

ObjectiveTo develop and validate an explainable artificial intelligence (AI) model for detecting geographic atrophy (GA) via colour retinal photographs.Methods and analysisWe conducted a prospective study where colour fundus images were...

Development of LuxIA, a Cloud-Based AI Diabetic Retinopathy Screening Tool Using a Single Color Fundus Image.

Diabetic retinopathy (DR) is the leading cause of vision impairment in working-age adults. Automated screening can increase DR detection at early stages at relatively low costs. We developed and evaluated a cloud-based screening tool that uses artificial intelligence (AI), the LuxIA algorithm, to detect DR from a single fundus image. Color fundus images that were previously graded by expert readers were collected from the Canarian Health Service (Retisalud) and used to train LuxIA, a deep-learning-based algorithm for the detection of more than mild DR. The algorithm was deployed in the Discovery cloud platform to evaluate each test set. Sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve were computed using a bootstrapping method to evaluate the algorithm performance and compared through different publicly available datasets. A usability test was performed to assess the integration into a clinical tool. Three separate datasets, Messidor-2, APTOS, and a holdout set from Retisalud were evaluated. Mean sensitivity and specificity with 95% confidence intervals (CIs) reached for these three datasets were 0.901 (0.901-0.902) and 0.955 (0.955-0.956), 0.995 (0.995-0.995) and 0.821 (0.821-0.823), and 0.911 (0.907-0.912) and 0.880 (0.879-0.880), respectively. The usability test confirmed the successful integration of LuxIA into Discovery. Clinical data were used to train the deep-learning-based algorithm LuxIA to an expert-level performance. The whole process (image uploading and analysis) was integrated into the cloud-based platform Discovery, allowing more patients to have access to expert-level screening tools. Using the cloud-based LuxIA tool as part of a screening program may give diabetic patients greater access to specialist-level decisions, without the need for consultation.

Segmentation of hard exudate lesions in color fundus image using two-stage CNN-based methods

The presence of hard exudate (EX) lesions is an early clinical symptom of Diabetic Retinopathy (DR); its accurate segmentation is essential for diagnosis and treatment. Automatic segmentation of EX lesions is challenging because they have a wide range of sizes, shapes, and brightness, and can be confused with soft exudate lesions. In this paper, we present and assess an efficient segmentation method for interactively segmenting EX lesions in fundus images from DR patients. Our approach consisted of two main stages: (1) the first stage generates an automatic segmentation using a CNN-based method. (2) If the obtained segmentation is suboptimal, the second stage is used to refine each inaccurately segmented region using a CNN-based interactive segmentation method. To train and evaluate method performance, we used two public datasets (IDRiD and DDR) and two local datasets from two medical centers (EHos and EWT) containing a wide range of EX lesions. Several state-of-the art CNN-based models and the proposed method were trained and evaluated on IDRiD, DDR and EHos datasets using the area under the precision–recall curve (AUPR), the dice similarity coefficient (Dice) and the intersection over union (IoU). In addition, EWT dataset was used to evaluate the methods without retraining the CNN models. The evaluation results using IDRiD dataset showed that the proposed method achieved AUPR, mean Dice and mean IoU scores of 0.893, 76.6% and 62.4%, respectively, which are comparable to the state-of-the-art method on EX segmentation. When evaluated on the DDR dataset with large portion of small EX lesions, the proposed method achieved AUPR, mean Dice and mean IoU scores of 0.693, 60.8% and 45.5%, yielding state-of-the-art performance. Furthermore, with the main evaluation metric being AUPR, the evaluation results on two datasets, EHos and ETW, with scores of 0.689 and 0.667 demonstrated superiority of the proposed method over the other segmentation methods. The proposed two-stage CNN-based segmentation method is an useful approach for accurately segmenting EX lesions in color fundus images, enabling lesion quantification to aid experts in treating patients with DR. The source code, the trained CNN models and the data are publicly available at: https://github.com/dvquang2000/Interactive_HardExudates_Segmentation.

Clinical Characteristics and Genetic Variants in Taiwanese Patients With PROM1-Related Inherited Retinal Disorders.

This study investigated the clinical characteristics of patients with PROM1-related inherited retinal diseases (IRDs). Patients diagnosed with IRDs who had mutations in PROM1 were identified at Linkou Chang Gung Memorial Hospital and Kaohsiung Medical University Hospital in Taiwan. Information on clinical characteristics and best-corrected visual acuity was recorded. Color fundus (CF) images, fundus autofluorescence photography (FAF), spectral-domain optical coherence tomography (SD-OCT), and electroretinograms (ERGs) were analyzed to examine patient phenotypes. PROM1 variants were detected using whole exome sequencing and verified by Sanger sequencing. Fourteen patients from nine families with PROM1-related IRDs were analyzed. Most patients exhibited chorioretinal atrophy in the macular area, with or without extramacular involvement on CF. Similarly, hypo-autofluorescence confined to the macular area, with or without extramacular involvement, was present for most patients on FAF. Furthermore, SD-OCT revealed outer retinal tubulations and focal or diffuse retinal thinning. ERGs showed variable findings, including maculopathy with normal ERG, subnormal cone response, and extinguished rod and cone responses. We detected five variants of the PROM1 gene, including c.139del, c.794del, c.1238T>A, c.2110C>T, and c.1117C>T. In this study, we evaluated 14 Taiwanese patients with five PROM1 variants. Additionally, incomplete penetrance of heterozygous PROM1 variants was observed. Furthermore, patients with autosomal dominant PROM1 variants had lesions in the macular area and the peripheral region of the retina. SD-OCT serves as a useful tool for early detection of PROM1-related IRDs, as it captures certain signs of such diseases.

A deep learning model for generating fundus autofluorescence images from color fundus photography

BackgroundFundus Autofluorescence (FAF) is a valuable imaging technique used to assess metabolic alterations in the retinal pigment epithelium (RPE) associated with various age-related and disease-related changes. The practical uses of FAF are ever-growing. This study aimed to evaluate the effectiveness of a generative deep learning (DL) model in translating color fundus (CF) images into synthetic FAF images and explore its potential for enhancing screening of age-related macular degeneration (AMD). MethodsA generative adversarial network (GAN) model was trained on pairs of CF and FAF images to generate synthetic FAF images. The quality of synthesized FAF images was assessed objectively by common generation metrics. Additionally, the clinical effectiveness of the generated FAF images in AMD classification was evaluated by measuring the area under the curve (AUC), using the LabelMe dataset. ResultsA total of 8410 FAF images from 2586 patients were analyzed. The synthesized FAF images exhibited an impressive objectively assessed quality, achieving a multi-scale structural similarity index (MS-SSIM) of 0.67. When evaluated on the LabelMe dataset, the combination of generated FAF images and CF images resulted in a noteworthy improvement in AMD classification accuracy, with the AUC increasing from 0.931 to 0.968. ConclusionsThis study presents the first attempt to use a generative deep learning model to create authentic and high-quality FAF images from CF images. The incorporation of the translated FAF images on top of CF images improved the accuracy of AMD classification. Overall, this study presents a promising approach to enhance large-scale AMD screening.

Energy efficiency in edge TPU vs. embedded GPU for computer-aided medical imaging segmentation and classification

In this work, we evaluate the energy usage of fully embedded medical diagnosis aids based on both segmentation and classification of medical images implemented on Edge TPU and embedded GPU processors. We use glaucoma diagnosis based on color fundus images as an example to show the possibility of performing segmentation and classification in real time on embedded boards and to highlight the different energy requirements of the studied implementations.Several other works develop the use of segmentation and feature extraction techniques to detect glaucoma, among many other pathologies, with deep neural networks. Memory limitations and low processing capabilities of embedded accelerated systems (EAS) limit their use for deep network-based system training. However, including specific acceleration hardware, such as NVIDIA's Maxwell GPU or Google's Edge TPU, enables them to perform inferences using complex pre-trained networks in very reasonable times.In this study, we evaluate the timing and energy performance of two EAS equipped with Machine Learning (ML) accelerators executing an example diagnostic tool developed in a previous work. For optic disc (OD) and cup (OC) segmentation, the obtained prediction times per image are under 29 and 43 ms using Edge TPUs and Maxwell GPUs, respectively. Prediction times for the classification subsystem are lower than 10 and 14 ms for Edge TPUs and Maxwell GPUs, respectively. Regarding energy usage, in approximate terms, for OD segmentation Edge TPUs and Maxwell GPUs use 38 and 190 mJ per image, respectively. For fundus classification, Edge TPUs and Maxwell GPUs use 45 and 70 mJ, respectively.

Joint DR-DME grading classification using optimal feature selection-based deep graph correlation network

Diabetic macular edema (DME) and diabetic retinopathy (DR) are the leading causes of human blindness, and accurate grading of individual DME, individual DR, and the joint DR-DME is very important for the diagnosis of human eye diseases. However, the conventional methods failed to separate the features such as disease-specific, disease-dependent, and joint DR-DME, which resulted in poor grading accuracy. In addition, optimal feature selection is also vital in DR-DME grading classification for improving the performance of joint DR-DME grading classification. Therefore, this work focuses on the implementation of an advanced deep graph correlation learning model based on a joint DR-DME network (JDD-Net) for disease detection and classification from color fundus images. Initially, convolutional block attention module (CBAM) and joint disease attention (JDA) modules are combined to extract the DR-specific, DME-specific, and joint DR-DME disease-dependent features. Here, interdependent DR and DME features are separated by a CBAM-based channel-spatial split attention mechanism. In addition, an iterative random forest network (IRF-Net) is used to select the optimal features by adopting fast machine learning properties. Finally, a deep graph correlation network (DGCN) is used to classify the different diseases using a pre-trained model. The simulations conducted on the Indian Diabetic Retinopathy Image Dataset (IDRiD) disclose that the proposed JDD-Net results in improved individual DR, individual DME, and joint DR-DME performance as compared to state-of-the-art approaches with DR, DME, and joint DR-DME accuracy of 99.53%, 99.1%, and 99.01%, respectively.