Search

Diabetic Retinopathy (DR) is one of the severe complications of diabetes which if not timely diagnosed and treated, can lead to severe retinal damage including irreversible vision loss. The overburdened healthcare system, lack of skilled ophthalmologists, and long delays in reporting the results of fundus images are the main reasons for untimely intervention. AI-based diagnostic systems can improve the situation. In fact, World Economic Forum has listed AI-facilitated healthcare as one of the leading emerging technologies. Many deep learning algorithms have been proposed for the diagnosis of DR in recent years but these studies only considered one or two symptoms for detecting diabetic retinopathy such as exudates, vessels, or hemorrhages, and not all of them together. In this paper, we propose a diagnostic system that takes into account all symptoms for detecting diabetic retinopathy and proposed a novel retinal vessels extraction algorithms that overcame the other studies vessel extraction algorithm. In this study, we also classify abnormal images as having exudates or not. For this purpose, fundus image datasets were annotated by experts for the presence or absence of exudates, these retinal images are then passed through pre-processing stage for extraction of haemorrhages and retinal vessels separately, these pre-processed images are then fused to one image and passed to two classifier, one classifier trained on normal and abnormal retinal images and second classifier trained to classify DR due to other symptoms (haemorrhages/neovascularization) and DR due to exudates. These classifiers are arranged in a cascaded way for identifying image with no DR, DR with exudates and DR with other symptoms. Results show that the proposed extraction algorithms improve the performance of the state-of-the-art DR classification model. Our model achieved 0.88 accuracy, with 0.91 precision, 0.86 recall, and 0.88 F1 score. [ABSTRACT FROM AUTHOR]

An eye disease affects the entire sensory operation, and an unrecognised and untreated eye disease may lead to loss of vision. The proposed work aims to develop an automated age-related macular degeneration (AMD) detection system using a Deep-Learning (DL) scheme with serially concatenated deep and handcrafted features. The research includes the following phases: initial data processing, deep-features extraction with VGG16, handcrafted feature extraction, optimal feature selection using Mayfly-Algorithm, serial features concatenation, and binary classification and validation. In this work, the handcrafted features, such as local binary pattern (LBP), pyramid histogram of oriented gradients (PHOG), and discrete wavelet transform (DWT), are extracted from the test images and concatenated with the deep-features of VGG16. The performance of the developed system is separately tested using fundus retinal images (FRI) and optical coherence tomography (OCT) images. A binary classification with a fivefold cross validation is employed and the best outcome of this trial is chosen as the result. The performance of VGG16 is then compared with that of VGG19, ResNet50, and AlexNet. The experimental outcome of this research confirms that the proposal of VGG16 with concatenated features achieves AMD detection accuracies of 97.08 and 97.50 for FRI and OCT images, respectively. [ABSTRACT FROM AUTHOR]

Abstract Background Ocular involvement may occur via several mechanisms in patients with transfusion-dependent β-thalassemia (TDT) mainly chronic anemia, iron overload and iron chelator toxicity. We aimed to evaluate the frequency of abnormal ocular findings and their relationship with hematologic parameters in TDT patients. Methods In this cross-sectional study from January 2018 to January 2019, a total of 79 patients with TDT over the age of 18 who were on iron-chelation therapy (ICT) were consecutively investigated. All patients were registered at the Thalassemia Comprehensive Center affiliated with Shiraz University of Medical Sciences, Shiraz, Southern Iran. Complete ophthalmic examination was performed by an expert ophthalmologist. Clinical and hematologic parameters were collected from the patients´ medical records. Results The mean age ± standard deviation (SD) of the patients was 28.4 ± 5.6 years (range: 18–43). Twenty-four patients (30.4%) were male and 29 (36.7%) were splenectomized. The mean ± SD of the best-corrected visual acuity (VA) was 0.960 ± 0.086 decimal, (range: 0.6–1), 0.016 ± 0.046 logMar, (range: 0–0.2). The frequency of patients with VA > 0.1 logMar was 3 (3.8%). The mean intraocular pressure (IOP) was 14.88 ± 3.34 (6–25) mmHg. Fundus abnormalities were observed in 8 patients (10.1%), consisting of increased cup-disk ratio (3.8%), vessel tortuosity (2.5%), retinal pigment epithelium degeneration (2.5%), myelinated nerve fiber layer (1.3%), and internal limiting membrane wrinkling (1.3%). No significant association was observed between fundus abnormalities, VA, or IOP with hematologic parameters (P > 0.05). TDT patients with diabetes mellitus had significantly higher IOP (P = 0.010) but similar frequency of fundus abnormalities with non-diabetic patients (P > 0.05). Conclusions The frequency of ocular abnormalities in our patients was lower than the previous reports. The frequency of fundus abnormalities were similar in diabetic and non-diabetic thalassemia patients indicating close monitoring and proper management of the disease and comorbidities in these patients.

INTRODUCTION: Macular edema is not a disease itself, but, a very common condition in most of the retinal diseases, such as diabetic retinopathy, retinal vein occlusion, hypertensive retinopathy, age-related macular edema, etc. and postocular surgery. OBJECTIVES: We have discussed how macular edema plays an important role in blindness in case of various retinal blood vascular diseases and post-ophthalmic surgery. We have analyzed vast state-of-the-art methods for retinal abnormality detection. METHODS: The proposed method uses a semi-automated macula segmentation approach and Local Binary Pattern features to train k-Nearest Neighbor classifier and performs binary classification. RESULTS: We have achieved 80% accuracy and 90% sensitivity in classifying normal and abnormal retina. CONCLUSION: We justified the notion that it will be beneficial to have a method that can analyse the macula region and alert if there is any abnormality near that region to prevent vision loss. [ABSTRACT FROM AUTHOR]

INTRODUCTION: Macular edema is not a disease itself, but, a very common condition in most of the retinal diseases, such as diabetic retinopathy, retinal vein occlusion, hypertensive retinopathy, age-related macular edema, etc. and post-ocular surgery. OBJECTIVES: We have discussed how macular edema plays an important role in blindness in case of various retinal blood vascular diseases and post-ophthalmic surgery. We have analyzed vast state-of-the-art methods for retinal abnormality detection.METHODS: The proposed method uses a semi-automated macula segmentation approach and Local Binary Pattern features to train k-Nearest Neighbor classifier and performs binary classification. RESULTS: We have achieved 80% accuracy and 90% sensitivity in classifying normal and abnormal retina.CONCLUSION: We justified the notion that it will be beneficial to have a method that can analyse the macula region and alert if there is any abnormality near that region to prevent vision loss.

Eye is an essential sensory organ in human physiological system and disease in eye severely affects the vision system. Age-related-macular-degeneration (AMD) is a common eye disease in elderly people (age > 60 years) and the untreated AMD will cause severe eye illness including permanent vision loss. This research work proposes a machine-learning-system (MLS) to detect the AMD using the fundus-retinal-images. The various phases of the MLS employed in the proposed work is as follows; (i) Image thresholding based on Shannon' Entropy and Bat Algorithm (BA + SE), (ii) Gaussian-filter (GF) based image smoothening, (iii) feature extraction, (iv) Feature selection using statistical test, and (v) classifier implementation and validation. The image features, such as the gray level co-occurrence matrix and entropies are considered to categorize the retinal image database into AMD/Non-AMD class. The performance of classifiers, such as Naïve–Bayes (NB), decision-tree (DT), K-nearest neighbors, random-forest (RF) and support-vector-machine with linear kernel (SVM) are compared and the experimental outcome confirms that, the classification accuracy of SVM is superior (> 93%) compared to other classifiers. [ABSTRACT FROM AUTHOR]

Background: Mucopolysaccharidosis type I (MPS I) is a lysosomal storage disorder caused by α‐L‐iduronidase deficiency, resulting in accumulation of glycosaminoglycans (GAG). Ophthalmological manifestations are common in MPS I patients and often lead to visual impairment. Accumulation of GAG in corneal or retinal tissues reduces vision causing corneal opacity and neurosensory complications. One available treatment for MPS I patients is enzyme replacement therapy (ERT), but the results of such treatment on eye disease are still debatable. Therefore, we aimed to determine the progression of ocular manifestations as well as the effectiveness of intravenous ERT in MPS I. Methods: Corneal and retinal analyses were perform in eyes from 2‐ to 8‐month normal and MPS I mice. Some MPS I mice received ERT (1.2 mg/kg of laronidase) every 2 weeks from 6 to 8 months and histological findings were compared with controls. Additionally, cornea from two MPS I patients under ERT were evaluated. Results: Mouse corneal tissues had GAG accumulation early in life. In the retina, we found a progressive loss of photoreceptor cells, starting at 6 months. ERT did not improve or stabilize the histological abnormalities. MPS I patients, despite being on ERT for over a decade, presented GAG accumulation in the cornea, corneal thickening, visual loss and needed corneal transplantation. Conclusion: We provide data on the time course of ocular alteration in MPS I mice. Our results also suggest that ERT is not effective in treating the progressive ocular manifestations in MPS I mice and fails to prevent corneal abnormalities in patients. [ABSTRACT FROM AUTHOR]

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that manifests as cognitive deficits and memory decline, especially in old age. Several biomarkers have been developed to monitor AD progression. Given that the retina and brain share some similarities including features related to anatomical composition and neurological functions, the retina is closely associated with the progression of AD. Herein, we review the evidence of retinal dysfunction in AD, particularly at the early stage, together with the underlying molecular mechanisms. Furthermore, we compared the retinal pathologies of AD and other ophthalmological diseases and summarized potential retinal biomarkers measurable by existing technologies for detecting AD, providing insights for the future development of diagnostic tools.

Alzheimer's disease (AD) is a progressive neurodegenerative disorder that manifests as cognitive deficits and memory decline, especially in old age. Several biomarkers have been developed to monitor AD progression. Given that the retina and brain share some similarities including features related to anatomical composition and neurological functions, the retina is closely associated with the progression of AD. Herein, we review the evidence of retinal dysfunction in AD, particularly at the early stage, together with the underlying molecular mechanisms. Furthermore, we compared the retinal pathologies of AD and other ophthalmological diseases and summarized potential retinal biomarkers measurable by existing technologies for detecting AD, providing insights for the future development of diagnostic tools. [ABSTRACT FROM AUTHOR]