Fundus Fluorescein Angiography Images Research Articles

FFA-GPT: an automated pipeline for fundus fluorescein angiography interpretation and question-answer

Fundus fluorescein angiography (FFA) is a crucial diagnostic tool for chorioretinal diseases, but its interpretation requires significant expertise and time. Prior studies have used Artificial Intelligence (AI)-based systems to assist FFA interpretation, but these systems lack user interaction and comprehensive evaluation by ophthalmologists. Here, we used large language models (LLMs) to develop an automated interpretation pipeline for both report generation and medical question-answering (QA) for FFA images. The pipeline comprises two parts: an image-text alignment module (Bootstrapping Language-Image Pre-training) for report generation and an LLM (Llama 2) for interactive QA. The model was developed using 654,343 FFA images with 9392 reports. It was evaluated both automatically, using language-based and classification-based metrics, and manually by three experienced ophthalmologists. The automatic evaluation of the generated reports demonstrated that the system can generate coherent and comprehensible free-text reports, achieving a BERTScore of 0.70 and F1 scores ranging from 0.64 to 0.82 for detecting top-5 retinal conditions. The manual evaluation revealed acceptable accuracy (68.3%, Kappa 0.746) and completeness (62.3%, Kappa 0.739) of the generated reports. The generated free-form answers were evaluated manually, with the majority meeting the ophthalmologists’ criteria (error-free: 70.7%, complete: 84.0%, harmless: 93.7%, satisfied: 65.3%, Kappa: 0.762–0.834). This study introduces an innovative framework that combines multi-modal transformers and LLMs, enhancing ophthalmic image interpretation, and facilitating interactive communications during medical consultation.

Unified deep learning model for predicting fundus fluorescein angiography image from fundus structure image

The prediction of fundus fluorescein angiography (FFA) images from fundus structural images is a cutting-edge research topic in ophthalmological image processing. Prediction comprises estimating FFA from fundus camera imaging, single-phase FFA from scanning laser ophthalmoscopy (SLO), and three-phase FFA also from SLO. Although many deep learning models are available, a single model can only perform one or two of these prediction tasks. To accomplish three prediction tasks using a unified method, we propose a unified deep learning model for predicting FFA images from fundus structure images using a supervised generative adversarial network. The three prediction tasks are processed as follows: data preparation, network training under FFA supervision, and FFA image prediction from fundus structure images on a test set. By comparing the FFA images predicted by our model, pix2pix, and CycleGAN, we demonstrate the remarkable progress achieved by our proposal. The high performance of our model is validated in terms of the peak signal-to-noise ratio, structural similarity index, and mean squared error.

Dual contrastive learning for synthesizing unpaired fundus fluorescein angiography from retinal fundus images.

Fundus fluorescein angiography (FFA) is an imaging method used to assess retinal vascular structures by injecting exogenous dye. FFA images provide complementary information to that provided by the widely used color fundus (CF) images. However, the injected dye can cause some adverse side effects, and the method is not suitable for all patients. To meet the demand for high-quality FFA images in the diagnosis of retinopathy without side effects to patients, this study proposed an unsupervised image synthesis framework based on dual contrastive learning that can synthesize FFA images from unpaired CF images by inferring the effective mappings and avoid the shortcoming of generating blurred pathological features caused by cycle-consistency in conventional approaches. By adding class activation mapping (CAM) to the adaptive layer-instance normalization (AdaLIN) function, the generated images are made more realistic. Additionally, the use of CAM improves the discriminative ability of the model. Further, the Coordinate Attention Block was used for better feature extraction, and it was compared with other attention mechanisms to demonstrate its effectiveness. The synthesized images were quantified by the Fréchet inception distance (FID), kernel inception distance (KID), and learned perceptual image patch similarity (LPIPS). The extensive experimental results showed the proposed approach achieved the best results with the lowest overall average FID of 50.490, the lowest overall average KID of 0.01529, and the lowest overall average LPIPS of 0.245 among all the approaches. When compared with several popular image synthesis approaches, our approach not only produced higher-quality FFA images with clearer vascular structures and pathological features, but also achieved the best FID, KID, and LPIPS scores in the quantitative evaluation.

Superselective Arterial Hyaluronidase Thrombolysis is an Effective Treatment for Hyaluronic Acid-Induced Retinal Artery Occlusion: Study in a Rabbit Model.

There are serious complications associated with hyaluronic acid (HA) facial injections, including vision impairment due to retinal artery ischemia. In this study, we put forth a clinically relevant model of retinal ischemia and reperfusion in rabbit. We used this to verify the efficacy of hyaluronidase intra-artery thrombolysis in the treatment of hyaluronic acid-induced retinal artery occlusion. Retinal artery ischemia was induced by injecting HA into the ophthalmic artery (OA) of adult chinchilla rabbit, and reperfusion was achieved by intra-artery thrombolysis therapy with hyaluronidase following 60min and 4h of occlusion. Digital subtraction angiography (DSA) and fundus fluorescein angiography (FFA) were used to evaluate blood flow in the retina. Electroretinogram (ERG), hematoxylin and eosin staining and transmission electron microscope were used to evaluate the structure and function of the retina after ischemia and reperfusion following 60min and 4h of occlusion. DSA and FFA images confirmed occlusion of the ophthalmic and central retinal arteries, as well as reperfusion after hyaluronidase thrombolysis. ERG indicated retinal dysfunction following ischemia, and thrombolysis partially rescued its impairment following 4h of occlusion. Hematoxylin and eosin staining and TUNEL staining revealed ischemia-induced histological damages in the retina at different time windows, and hyaluronidase thrombolysis partially mitigated these damages. We report a method to establish a HA-induced retinal artery occlusion animal model. Hyaluronidase intra-artery thrombolysis was used to recanalize the embolized OA at different time points. Using our method, we achieved retinal reperfusion, and an improvement was observed in the visual function of rabbits after hyaluronidase thrombolysis following 4h of occlusion. We believe that hyaluronidase intra-artery thrombolysis is an effective method to treat HA-induced retinal artery occlusion in clinic. This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

Comparison of the Efficiency and Safety of Photodynamic Therapy Combined Intravitreal Bevacizumab and Intravitreal Triamcinolone Treatments in Age-Related Macular Degeneration

Objective: This study aims to retrospectively examine the mid and long-term results of the photodynamic therapy (PDT) +IVTA group and the PDT +IVB applied Group in the treatment of choroid neovascularization (CNV) due to age-related macular degeneration (AMD) and to compare the results with the results of multicenter clinical studies. Materials and Methods: 64 eyes of 64 patients admitted to Haseki Training and Research Hospital, Department of Ophthalmology, Retina unit between June 2003 and March 2009, diagnosed with CNV secondary to age-related macular degeneration and treated, were included in the study. Thirty-five eyes with PDT +IVTA were classified as Group I, and 29 with PDT +IVTA were classified as Group II. Visual acuities of the cases were evaluated with ETDR and Snellen charts and converted to their equivalents on the Logarithm of Minimal Angle of Resolution (logMAR) scale. All patients underwent an ophthalmological examination before and after the treatment; fundus fluorescein angiography (FFA) and optical coherence tomography (OCT) were performed to see if there was any recurrence in the patients. Patients who were found to have relapsed were re-treated. Results: There were 26 (40.6%) female cases and 38 (59.4%) male cases. The ages ranged from 55 to 88, with a mean age of 73.85±7.26. The follow-up period of the patients ranged from 5 months to 69 months; the mean was 17.65±13.48 months. When evaluated according to FFA images, 40 (62.5%) were occult, 12 (18.8%) were minimal-classical, and 12 (18.8%) were dominant-classical. After treatment, visual acuity decreased in 20% of the PDT +IVTA group and increased or remained the same in 80%. In the PDT +IVB group, blurred vision decreased by 27.6% and increased or remained the same by 72.4%. There was no statistically significant difference between the groups regarding the change in visual acuity after treatment compared to before treatment (p>0.05). In OCT, the decrease in the mean central foveal thickness (CFT) after treatment compared to the mean CFT before treatment in Group I was statistically significant (p<0.05). In Group II - the decrease in the mean CFT after treatment compared to the mean CFT before treatment was statistically significant (p<0.01). In Group I, there was no statistically significant change in the mean lesion diameter after treatment compared to the mean lesion diameter before treatment (p>0.05). However, the decrease in Group II was statistically significant (p<0.01). While the increase in CFT (central foveal thickness) and lesion diameter was more pronounced in the PDT+IVTA group post-treatment compared to the pre-treatment group, no significant difference was observed in visual acuity change within the PDT+IVTA group. Conclusion: In the treatment of choroidal neovascularization secondary to age-related macular degeneration, the combination of Triamcinolone or Bevacizumab with photodynamic therapy has proven to be both effective and safe. This approach helps preserve the current vision and even enhance visual outcomes. Combination therapy reduces the frequency and number of monotherapies used to achieve visual success.

Multimodal Imaging of Reticular Pseudodrusen in Turkish Patients.

To investigate the presence and prevalence of reticular pseudodrusen (RPD) in patients with age-related macular degeneration using multiple imaging modalities and to compare the sensitivity and specificity of these modalities in the detection of RPD. Images from a total of 198 consecutive patients were analyzed prospectively. Color fundus photography, red-free imaging, spectral domain optical coherence tomography (SD-OCT), infrared and blue reflectance (BR) imaging, fundus autofluorescence (FAF), enhanced-depth imaging OCT (EDI-OCT), fundus fluorescein angiography (FFA) and indocyanine green angiography were performed. RPD was diagnosed in the presence of relevant findings in at least two of the imaging methods used. RPD were detected in 149 eyes (37.6%). In the detection of RPD, color fundus photography, red-free photography, SD-OCT, infrared, FAF, BR, and FFA imaging had sensitivity values of 50%, 57.7%, 91.6%, 95%, 74.6%, 65.7%, and 28.2% and specificity values of 99.6%, 100%, 98.4%, 94.6%, 100%, 99.6%, and 69.8%, respectively. Infrared imaging had the highest sensitivity. SD-OCT combined with infrared imaging was the most sensitive imaging technique for detecting RPD. The high specificity of FAF, red-free, and BR imaging may be useful to confirm a diagnosis of RPD.

An artificial intelligence system for the whole process from diagnosis to treatment suggestion of ischemic retinal diseases

Ischemic retinal diseases (IRDs) are a series of common blinding diseases that depend on accurate fundus fluorescein angiography (FFA) image interpretation for diagnosis and treatment. An artificial intelligence system (Ai-Doctor) was developed to interpret FFA images. Ai-Doctor performed well in image phase identification (area under the curve [AUC], 0.991-0.999, range), diabetic retinopathy (DR) and branch retinal vein occlusion (BRVO) diagnosis (AUC, 0.979-0.992), and non-perfusion area segmentation (Dice similarity coefficient [DSC], 89.7%-90.1%) and quantification. The segmentation model was expanded to unencountered IRDs (central RVO and retinal vasculitis), with DSCs of 89.2% and 83.6%, respectively. A clinically applicable ischemia index (CAII) was proposed to evaluate ischemic degree; patients with CAII values exceeding 0.17 in BRVO and 0.08 in DR may be associated with increased possibility for laser therapy. Ai-Doctor is expected to achieve accurate FFA image interpretation for IRDs, potentially reducing the reliance on retinal specialists.

Translation of Color Fundus Photography into Fluorescein Angiography Using Deep Learning for Enhanced Diabetic Retinopathy Screening

To develop and validate a deep learning model that can transform color fundus (CF) photography into corresponding venous and late-phase fundus fluorescein angiography (FFA) images. Cross-sectional study. We included 51 370 CF-venous FFA pairs and 14 644 CF-late FFA pairs from 4438 patients for model development. External testing involved 50 eyes with CF-FFA pairs and 2 public datasets for diabetic retinopathy (DR) classification, with 86 952 CF from EyePACs, and 1744 CF from MESSIDOR2. We trained a deep-learning model to transform CF into corresponding venous and late-phase FFA images. The translated FFA images' quality was evaluated quantitatively on the internal test set and subjectively on 100 eyes with CF-FFA paired images (50 from external), based on the realisticity of the global image, anatomical landmarks (macula, optic disc, and vessels), and lesions. Moreover, we validated the clinical utility of the translated FFA for classifying 5-class DR and diabetic macular edema (DME) in the EyePACs and MESSIDOR2 datasets. Image generation was quantitatively assessed by structural similarity measures (SSIM), and subjectively by 2 clinical experts on a 5-point scale (1 refers real FFA); intragrader agreement was assessed by kappa. The DR classification accuracy was assessed by area under the receiver operating characteristic curve. The SSIM of the translated FFA images were > 0.6, and the subjective quality scores ranged from 1.37 to 2.60. Both experts reported similar quality scores with substantial agreement (all kappas > 0.8). Adding the generated FFA on top of CF improved DR classification in the EyePACs and MESSIDOR2 datasets, with the area under the receiver operating characteristic curve increased from 0.912 to 0.939 on the EyePACs dataset and from 0.952 to 0.972 on the MESSIDOR2 dataset. The DME area under the receiver operating characteristic curve also increased from 0.927 to 0.974 in the MESSIDOR2 dataset. Our CF-to-FFA framework produced realistic FFA images. Moreover, adding the translated FFA images on top of CF improved the accuracy of DR screening. These results suggest that CF-to-FFA translation could be used as a surrogate method when FFA examination is not feasible and as a simple add-on to improve DR screening. Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Diabetic microaneurysms detected by fluorescein angiography spatially correlate with regions of macular ischemia delineated by optical coherence tomography angiography.

To characterize the relationship between diabetic macular ischemia (DMI) delineated by optical coherence tomography angiography (OCTA) and microaneurysms (MAs) identified by fundus fluorescein angiography (FFA). Patients with diabetic retinopathy (DR) who underwent OCTA and FFA were retrospectively identified. FFA images were cropped and aligned with their respective OCTA images using i2k Align Retina software (Dual-Align, Clifton Park, NY, USA). Foveal avascular zone (FAZ) and ischemic areas were manually delineated on OCTA images, and MAs were marked on the corresponding FFA images before overlaying paired scans for analysis (ImageJ; National Institutes of Health, Bethesda, MD, USA). Twenty-eight eyes of 20 patients were included. The average number of MAs identified in cropped FFA images was 127 ± 42. More DMI was noted in the superficial capillary plexus (SCP; 36 ± 13%) compared to the deep capillary plexus (DCP; 28 ± 14%, P < 0.001). Similarly, more MAs were associated with ischemic areas in SCP compared to DCP (92.0 ± 35.0 vs. 76.8 ± 36.5, P < 0.001). Most MAs bordered ischemic areas; fewer than 10% localized inside these regions. As DMI area increased, so did associated MAs (SCP: r = 0.695, P < 0.001; DCP: r = 0.726, P < 0.001). Density of MAs surrounding FAZ (7.7 ± 6.0 MAs/mm2) was similar to other DMI areas (SCP: 7.0 ± 4.0 MAs/mm2, P = 0.478; DCP: 9.2 ± 10.9 MAs/mm2, P = 0.394). MAs identified in FFA strongly associate with, and border areas of, DMI delineated by OCTA. Although more MAs are localized to SCP ischemia, the concentration of MAs associated with DCP ischemia is greater. By contrast, few MAs are present inside low-flow regions, likely because capillary loss is associated with their regression.

FDXR-associated disease in a Chinese cohort: Unraveling expanded ocular phenotypes and genetic spectrum

FDXRassociated disease is characterized by optic atrophy, acoustic neuropathy, and developmental delays. This study evaluated the ocular phenotypes and genetic features of patients with biallelic FDXR variants. Five individuals from unrelated non-consanguineous Chinese families with biallelic FDXR variants were identified using whole exome sequencing, Sanger sequencing, and co-segregation validation. In addition to optic atrophy and diverse extraocular manifestations, all patients presented with retinal dystrophy, and electroretinogram showed severely impaired cone and rod functions in their first decades. Three of the five patients showed attenuated retinal vessels that appeared as white lines on the fundus, and fundus fluorescein angiography (FFA) further revealed vascular abnormalities including delayed filling, completely occluded retinal vasculature, and severe retinal vascular nonperfusion of the peripheral retina. Five novel FDXR variants were identified: c.383C > T (p.A128V), c.963delG (p.R322fs*7), c.1052_1053delTC (p.L351Pfs*12), c.394-11T > G and c.1002+1G > A. Retinal dystrophy with attenuated retinal vessels appearing as white lines was observed in this cohort, and the FFA images revealed that retinal vascular occlusion could be a distinct clinical characteristic of FDXR-associated disease. Probands with FDXR revealed severe early onset ophthalmic features with rapid-progression, indicating the importance of early diagnosis and treatment. Moreover, this is the first study to report FFA manifestations in an FDXR cohort, expanding the FDXR-associated ocular disease phenotype and genetic spectrum.

One-shot Retinal Artery and Vein Segmentation via Cross-modality Pretraining

To perform one-shot retinal artery and vein segmentation with cross-modality artery-vein (AV) soft-label pretraining. Cross-sectional study. The study included 6479 color fundus photography (CFP) and arterial-venous fundus fluorescein angiography (FFA) pairs from 1964 participants for pretraining and 6 AV segmentation data sets with various image sources, including RITE, HRF, LES-AV, AV-WIDE, PortableAV, and DRSplusAV for one-shot finetuning and testing. We structurally matched the arterial and venous phase of FFA with CFP, the AV soft labels were automatically generated by utilizing the fluorescein intensity difference of the arterial and venous-phase FFA images, and the soft labels were then used to train a generative adversarial network to learn to generate AV soft segmentations using CFP images as input. We then finetuned the pretrained model to perform AV segmentation using only one image from each of the AV segmentation data sets and test on the remainder. To investigate the effect and reliability of one-shot finetuning, we conducted experiments without finetuning and by finetuning the pretrained model on an iteratively different single image for each data set under the same experimental setting and tested the models on the remaining images. The AV segmentation was assessed by area under the receiver operating characteristic curve (AUC), accuracy, Dice score, sensitivity, and specificity. After the FFA-AV soft label pretraining, our method required only one exemplar image from each camera or modality and achieved similar performance with full-data training, with AUC ranging from 0.901 to 0.971, accuracy from 0.959 to 0.980, Dice score from 0.585 to 0.773, sensitivity from 0.574 to 0.763, and specificity from 0.981 to 0.991. Compared with no finetuning, the segmentation performance improved after one-shot finetuning. When finetuned on different images in each data set, the standard deviation of the segmentation results across models ranged from 0.001 to 0.10. This study presents the first one-shot approach to retinal artery and vein segmentation. The proposed labeling method is time-saving and efficient, demonstrating a promising direction for retinal-vessel segmentation and enabling the potential for widespread application. Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Mid-phase pinpoint hyperfluorescent spots on fundus fluorescein angiography in acute central retinal artery occlusion – a novel imaging finding

PurposeTo describe the development and outcome of mid-phase pinpoint hyperfluorescent spots (MPHS) on fundus fluorescein angiography (FFA) in acute (< 7-day symptom onset) central retinal artery occlusion (CRAO) patients.MethodsThis retrospective study included acute CRAO patients who underwent multimodal imaging utilizing optical coherence tomography (OCT) and FFA between June 2017 and January 2023. The correlation of FFA images with the OCT images in various stages and severity of CRAO were studied.ResultsTwenty-three FFA studies on 23 patients with acute CRAO were included. In 11 (48%) cases, an important finding noted on FFA was the presence of single or multiple MPHS with adjacent minimal late vascular leakage. Of these 11 cases, eight (73%) were males and median age was 41 years (inter quartile range: 33–55 years). Visual acuity ranged from ‘light perception’ to 6/18, and these patients presented anytime on the same day to seven days after symptom onset. On OCT, three eyes had severe CRAO, seven eyes had moderate CRAO, and one eye had mild CRAO. MPHS were primarily observed at the posterior pole and more frequently observed in moderate CRAO severity. During follow-up, the MPHS and retinal vessel staining on FFA disappeared as the CRAO showed signs of resolution.ConclusionMPHS at the posterior pole on FFA in acute CRAO patients could indicate a more severe occlusion and poor visual outcomes, even after treatment. This finding is most likely caused by red blood cell aggregation or rouleaux formation.Trial registration numberNot applicable.

Neovascular Glaucoma Associated with Chronic Rhegmatogenous Retinal Detachment.

To demonstrate the clinical features and natural course of chronic retinal detachment-associated neovascular glaucoma. Ten patients, diagnosed with chronic retinal detachment-associated neovascular glaucoma during 2007-2016 were retrospectively investigated. Besides chronic retinal detachment, no patients had any neovascular glaucoma-predisposing conditions, such as carotid artery disease. Retinal perfusion status was evaluated from the fundus fluorescein angiography images. The mean age of patients was 57.5 years (range, 22-78 years). Complete retinal reattachment was achieved in three eyes, while partial or total chronic retinal detachment persisted in seven eyes. Wide-angle fundus fluorescein angiography revealed peripheral retinal capillary obstruction and severe nonperfusion. Neovascular glaucoma developed 213.4 months (range, 17-634 months) after retinal detachment. Three eyes received Ahmed valve implantation, while five eyes received intravitreal bevacizumab injection. Intraocular pressure was controlled in 10 eyes. Two eyes developed phthisis bulbi during follow-up. In eyes with a chronic retinal detachment history, iris neovascularization and neovascular glaucoma can develop due to retinal capillary obstruction and chronic retinal ischemia, even after achieving retinal reattachment. We suggest regular follow-up examinations for patients with chronic retinal detachment, particularly for eyes with retinal nonperfusion, as detected on fundus fluorescein angiography.

Model long-range dependencies for multi-modality and multi-view retinopathy diagnosis through transformers

Early eye examination based on fundus images effectively prevents visual impairment caused by retinopathy. The laborious and error-prone process of interpreting fundus images and the lack of ophthalmologists have driven research toward automated retinopathy diagnosis. However, most previous studies have focused on single-modality fundus images, disregarding the integration of information from multiple views, rendering the results unsatisfactory and inconsistent with clinical practice due to the incomprehensive lesion features and incomplete fundus fields. To address this issue, we introduce multi-modality and multi-view fundus images into the automated retinopathy diagnosis pipeline. In contrast to single fundus images, sequential relationships in multi-modality and multi-view fundus images contain essential long-range dependency information, which is vital for retinopathy diagnosis. Inspired by the recent success of transformers for excavating long-range dependencies in sequence data, in this paper, we propose a transformer-based automated retinopathy diagnosis framework for pathology classification and symptom report generation by integrating multi-modality and multi-view fundus images. Specifically, we present two transformer-based networks to construct long-range dependencies in different fundus images. Moreover, we adopt two novel modules to aggregate features of different modalities and views by modeling long-range dependencies among different fundus image sequences. Experiments are conducted on two in-house datasets, in which each subject provides one color fundus photography image and four-view fundus fluorescein angiography images. The experimental results of retinopathy classification and report generation tasks indicate that our proposed method is superior to other benchmarking methods, achieving a classification accuracy of 85.49% and a report generation BlEU-1 of 0.422.

Optic disc and retinal vascular features in first 6 years of Chinese children.

Retinal microvasculature plays an important role in children's fundus lesions and even in their later life. However, little was known on the features of normal retina in early life. The purpose of this study was to explore the normal retinal features in the first 6 years of life and provide information for future research. Children, aged from birth to 6 years old and diagnosed with various unilateral ocular diseases were included. Venous phase fundus fluorescein angiography images with the optic disc at the center were collected. Based on the ResUNet convolutional neural network, optic disc and retinal vascular features in the posterior retina were computed automatically. A total of 146 normal eyes of 146 children were included. Among different age groups, no changes were shown in the optic disc diameter (y = -0.00002x + 1.362, R2 = 0.025, p = 0.058). Retinal vessel density and fractal dimension are linearly and strongly correlated (r = 0.979, p < 0.001). Older children had smaller value of fractal dimension (y = -0.000026x + 1.549, R2 = 0.075, p = 0.001) and narrower vascular caliber if they were less than 3 years old (y = -0.008x + 84.861, R2 = 0.205, p < 0.001). No differences were in the density (y = -0.000007x + 0.134, R2 = 0.023, p = 0.067) and the curvature of retinal vessels (lnC = -0.00001x - 4.657, R2 = 0.001, p = 0.667). Age and gender did not impact the optic disc diameter, vessel density, and vessel curvature significantly in this group of children. Trends of decreased vessel caliber in the first 3 years of life and decreased vessel complexity with age were observed. The structural characteristics provide information for future research to better understand the developmental origin of the healthy and diseased retina.

Semi-Supervised Dual Stream Segmentation Network for Fundus Lesion Segmentation.

Accurate segmentation of retinal images can assist ophthalmologists to determine the degree of retinopathy and diagnose other systemic diseases. However, the structure of the retina is complex, and different anatomical structures often affect the segmentation of fundus lesions. In this paper, a new segmentation strategy called a dual stream segmentation network embedded into a conditional generative adversarial network is proposed to improve the accuracy of retinal lesion segmentation. First, a dual stream encoder is proposed to utilize the capabilities of two different networks and extract more feature information. Second, a multiple level fuse block is proposed to decode the richer and more effective features from the two different parallel encoders. Third, the proposed network is further trained in a semi-supervised adversarial manner to leverage from labeled images and unlabeled images with high confident pseudo labels, which are selected by the dual stream Bayesian segmentation network. An annotation discriminator is further proposed to reduce the negativity that prediction tends to become increasingly similar to the inaccurate predictions of unlabeled images. The proposed method is cross-validated in 384 clinical fundus fluorescein angiography images and 1040 optical coherence tomography images. Compared to state-of-the-art methods, the proposed method can achieve better segmentation of retinal capillary non-perfusion region and choroidal neovascularization.

A new hybrid approach for automated leakage detection in fundus fluorescein angiography images

A new hybrid approach for automated leakage detection in fundus fluorescein angiography images

Lesion-aware generative adversarial networks for color fundus image to fundus fluorescein angiography translation

Background and Objective: Fundus fluorescein angiography (FFA) is widely used in clinical ophthalmic diagnosis and treatment with the requirement of adverse fluorescent dyes injection. Recently, many deep Convolutional Neural Network(CNN)-based methods have been proposed to estimate FFA from color fundus (CF) images to eliminate the use of adverse fluorescent dyes. However, the robustness of these methods is affected by pathological changes.Method: In this work, we present a CNN-based approach, lesion-aware generative adversarial networks (LA-GAN), to enhance the visual effect of lesion characteristics in the generated FFA images. First, we lead the generator notice lesion information by joint learning with lesion region segmentation. A new hierarchical correlation multi-task framework for high-resolution images is designed. Second, to enhance the visual contrast between normal regions and lesion regions, a newly designed region-level adversarial loss is used rather than the image-level adversarial loss. The code is publicly available at: https://github.com/nicetomeetu21/LA-GAN.Results: The effectiveness of LA-Net has been verified in data with branch retinal vein occlusion. The proposed model reported as measures of generation performance a mean structural similarity (SSIM) of 0.536, mean learned perceptual image patch similarity (LPIPS) 0.312, outperforming other FFA generation and general image generation methods. Further, due to the proposed multi-task learning framework, the lesion-region segmentation performance was further reported as the mean Dice increased from 0.714 to 0.797 and the mean accuracy increased from 0.873 to 0.905, outperforming general single-task image segmentation methods.Conclusions: The results show that the visual effect of lesion characteristics can be improved by employing the region-level adversarial loss and the hierarchical correlation multi-task framework respectively. Based on the results of comparison with the state-of-the-art methods, LA-GAN is not only effective for CF-to-FFA translation, but also effective for lesion-region segmentation. Thus, it may be used for various image translation and lesion segmentation tasks in future research.

An automated diabetic retinopathy disorders detection model based on pretrained MobileNetv2 and nested patch division using fundus images

Aim: Fundus images are very important to diagnose some ophthalmologic disorders. Hence, fundus images have become a very important data source for machine-learning society. Our primary goal is to propose a new automated disorder classification model for diabetic retinopathy (DR) using the strength of deep learning. In this model, our proposed model suggests a treatment technique using fundus images. Material and Method: In this research, a new dataset was acquired and this dataset contains 1365 Fundus Fluorescein Angiography images with five classes. To detect these disorders automatically, we proposed a transfer learning-based feature engineering model. This feature engineering model uses pretrained MobileNetv2 and nested patch division to extract deep and exemplar features. The neighborhood component analysis (NCA) feature selection function has been applied to choose the top features. k nearest neighbors (kNN) classification function has been used to get results and we used 10-fold cross-validation (CV) to validate the results. Results: The proposed MobileNetv2 and nested patch-based image classification model attained 87.40% classification accuracy on the collected dataset. Conclusions: The calculated 87.40% classification accuracy for five classes has been demonstrated high classification accuracy of the proposed deep feature engineering model

Series-Parallel Generative Adversarial Network Architecture for Translating from Fundus Structure Image to Fluorescence Angiography

Although fundus fluorescein angiography (FFA) is a very effective retinal imaging tool for ophthalmic diagnosis, the requirement of intravenous injection of harmful fluorescein dye limits its application. As a screening diagnostic method that reduces the frequency of intravenous injection, a series-parallel generative adversarial network (GAN) architecture for translating fundus structure image to FFA images is proposed herein, using deep learning-based software that only needs an intravenous injection for the training process. Firstly, the fundus structure image and the corresponding FFA images of three phases are collected. Secondly, our series-parallel GAN is trained to translate FFA images from fundus structure image with the supervision of FFA images. Thirdly, the trained series-parallel GAN model is used to translate FFA images by only using fundus structure image. By comparing the FFA images translated by our algorithm, Sequence GAN, pix2pix, and cycleGAN, we show the advancement of our algorithm. To further confirm the advancements of our algorithm, we evaluate the peak signal-to-noise ratio (PSNR), structural similarity (SSIM) index, and mean-squared error (MSE) of our algorithm, Sequence GAN, pix2pix, and cycleGAN. To demonstrate the performance of our method, we show some typical FFA images translated by our algorithm.