Eye Images Research Articles

Vitreoretinal interface abnormalities in fellow eyes of patients with a macular hole.

This study aimed to determine the incidence of bilateral macular holes (MHs), and to investigate the incidence and characteristics of retinal abnormalities in the fellow eyes, using high-definition optical coherence tomography (OCT). Medical records from 724 consecutive patients diagnosed with full-thickness MHs, who underwent surgery at Fujita Health University Hospital from May 2015 to March 2022, were retrospectively reviewed. Based on the exclusion criteria, 658 patients with idiopathic MHs were included. B-scan OCT images of the fellow eyes were used to identify vitreoretinal interface abnormalities, such as MHs, epiretinal membranes (ERMs), vitreomacular traction syndrome (VMTS) and lamellar macular holes (LMHs). The mean age of the patients was 65.9±9.0 years. There were 292 males (44.4%) and 366 females (55.6%). There were 39 fellow eyes with MHs (5.9%), 77 with ERMs (11.7%), 32 with VMTS (4.9%) and 11 with LMHs (1.7%). Thirty-four fellow eyes (5.2%) could not be classified and 469 fellow eyes were normal (71.3%). Furthermore, the 39 bilateral MHs were divided into four stages according to the Gass classification, as follows: stage 1, 32 eyes (82.1%); stage 2, none (0%); stage 3, 5 eyes (12.8%); and stage 4, 2 eyes (5.1%). Vitreoretinal interface abnormalities, such as MHs, ERMs, VMTS and LMHs, are more common in the fellow eyes of patients with MHs than in the healthy population. The fellow eyes of patients with MHs should be carefully followed using OCT.

Comparing the clinical applicability of wavefront phase imaging in keratoconus versus normal eyes

The aim of this work is to quantitatively assess the wavefront phase of keratoconic eyes measured by the ocular aberrometer t·eyede (based on WaveFront Phase Imaging Sensor), characterized by a lateral resolution of 8.6 µm without requiring any optical element to sample the wavefront information. We evaluated the parameters: root mean square error, Peak-to-Valley, and amplitude of the predominant frequency (Fourier Transform analysis) of a section of the High-Pass filter map in keratoconic and healthy cohorts. Furthermore, we have analyzed keratoconic eyes that presented dark–light bands in this map to assess their period and orientation with the Fourier Transform. There are significant statistical differences (p value < 0.001) between healthy and keratoconic eyes in the three parameters, demonstrating a tendency to increase with the severity of the disease. Otherwise, the quantification of the bands reveals that the width is independent of eye laterality and keratoconic stage as orientation, which tends to be oblique. In conclusion, the quantitative results obtained with t·eyede could help to diagnose and monitor the progression of keratoconus.

Fusion of Thresholding and Limbic Pupilliary Boundary Algorithms for Iris Segmentation Process

Iris segmentation is crucial in the development of an effective iris recognition system, it involves localizing relatively the exact region of the iris before its features are being extracted for matching. Several researches have used different algorithms for the segmentation of iris, however, traditional systems encounter challenges in accurately identifying individuals under varying conditions, primarily due to inaccurate segmentation. To address this, the study proposes an AI-based fusion technique combining Thresholding and Limbic Pupillary Boundary algorithms, aiming to enhance segmentation accuracy and reliability. Leveraging the eye images datasets from reputable sources and MATLAB implementation, the study improves segmentation performance, particularly in uncertain environments. The results are presented using histogram representation to demonstrate the efficacy of the fusion approach, contributing to advancements in iris recognition technology.

A Lightweight Semi-supervised Eye Semantic Segmentation Method Based on Iterative Learning

Eye semantic segmentation has a significant role in promoting biometric recognition. Most works improve the accuracy through designing complicated feature learning modules that usually require a high computation cost and are prone to over-fit. In this work, we propose a lightweight eye semantic segmentation method based on iterative learning, which only requires a small amount of labeled eye image data under visible light conditions. To be specific, we first propose a pseudo-label filtering strategy for eye semantic segmentation, which leverages a semi-supervised incremental learning technique to solve the problem of label scarcity. Besides, an attention-guided lightweight high-representation module is designed to boost precision while ensuring efficiency. Experiments on the Ubiris dataset show that the proposed method achieves state-of-the-art performance, with a mIoU of 86.18%.

Classification of Diabetic Retinopathy using Deep Learning

Abstract: Diabetic retinopathy, a potentially fatal retinal disease, occurs in individuals with diabetes, causing vision damage and possible visual impairment traditionally, screening for diabetic retinopathy has been the norm hard by ophthalmologists manually. To facilitate this, we have DR. Our pre-trained model is trained on a large dataset of about 3662 training images, enabling it to automatically detect the DR platform in high-resolution fundus images The dataset used for this is publicly available on kaggle. The DR phases are categorized as 0, 1, 2, 3, and 4. The fundus eye images serve as the input parameters of the model. The pre-trained model then identifies the point of interest in bank eye images, then provides activation function insights By calculating weights, for example patients’ visual intensity levels, it helps classify diabetic retinopathy images into their intensity groups appropriately within the possibility of test- methods Provides variation.

‘BLUES’ procedure for assessing the blue level of the sclera in Osteogenesis Imperfecta

PurposeBlue sclera is a characteristic and common clinical sign of Osteogenesis Imperfecta (OI). However, there is currently no widely accepted, objective method for assessing and grading blue sclera in individuals with OI.To address this medical need, this study is aimed to design and validate a new method called ‘BLUES’ (BLUe Eye Sclera) to objectively identify and quantify the blue color in the sclera of patients affected by OI.MethodsSixty-two patients affected by OI and 35 healthy controls were enrolled in the present prospective study, for a total of 194 eyes analyzed.In the 'BLUES' procedure, eye images from patients with OI and control subjects were analyzed to assess and grade the blue level of the sclera using Adobe Photoshop Software. The validation process then involved comparing the results obtained with the ‘BLUES’ procedure to the judgement of experienced ophthalmologists (JEO).A receiver-operating characteristic (ROC) curve analysis was used to examine the overall discriminatory power. The sensitivity and specificity levels and the Cohen's Kappa (K) indexes of ‘BLUES’ and ‘JEO’ were estimated versus the standard OI diagnosis. The K indexes of ‘BLUES’ versus ‘JEO’ were also evaluated.ResultsThe optimal cut-off point of the scleral blue peak was calculated at 17%. Our findings demonstrated a sensitivity of 89% (CI95%: 0.835–0.945) and specificity of 87% (CI95%: 0.791–0.949) for the ‘BLUES’ procedure with an agreement versus the diagnosis of OI of 0.747. In comparison, the sensitivity and specificity of ‘JEO’ ranged from 89 to 94% and 77% to 100%, respectively, with an agreement ranging from 0.663 to 0.871 with the diagnosis of OI. The agreement between ‘BLUES ‘and ‘JEO’ evaluations ranged from 0.613 to 0.734.ConclusionsOur findings demonstrated an 89% sensitivity and an impressive 87% specificity of our method to analyze the blue sclera in OI. The results indicated high agreement with disease diagnosis and were consistent with evaluations by experienced ophthalmologists. The ‘BLUES’ procedure appears to be a simple, reliable and objective method for effectively identify and quantify the blue color of the sclera in OI.

Vitamin Deficiency Detection from eye Using Machine Learning

Vitamin deficiencies affect millions globally, with over two billion people at risk. Early detection is crucial, yet one in three youngsters lacks proper vitamin intake. Addressing this, a novel machine learning-based approach, using eye images instead of blood samples, offers a free desktop application for vitamin deficiency detection. The software, trained to distinguish healthy and deficient eyes, provides users with comprehensive reports. Early identification of deficiencies can prevent anemia, infectious illnesses, and developmental issues. Real-world trials confirm the method's superior efficiency compared to previous approaches, marking a promising advancement in global health. Keywords: Machine Learning, Opthalmology, YOLO, Deep Learning, Image preprocessing , extraction, ultralytics.

Comparison of Blue-Light Autofluorescence and Ultrawidefield Green-Light Autofluorescence for Assessing Geographic Atrophy

The goal of this study was to evaluate and compare the intermodality and interreader agreement of manual and semiautomated geographic atrophy (GA) area measurements in eyes with GA due to age-related macular degeneration (AMD) using conventional blue-light fundus autofluorescence (FAF) and ultrawidefield (UWF) green-light FAF systems. Prospective Cohort Study. Seventy-two eyes of 50 patients with a diagnosis of advanced nonneovascular AMD with GA. Fundus autofluorescence images of eyes with GA were obtained during a single visit using both the Spectralis HRA+ OCT2 device and the Optos California device. The area of the GA lesion(s) was segmented and quantified (mm2) with a fully manual approach where the lesions were outlined using Optos Advance and Heidelberg Eye Explorer (HEYEX) software. In addition, for the Heidelberg blue FAF images, GA lesions were also measured using the instrument's semiautomated software (Region Finder 2.6.4). For comparison between modalities/grading method, the mean values of the 2 graders were used. Intraclass correlation coefficients were computed to judge the agreement between graders. Seventy-two eyes of 50 patients were included in this study. There was nearly perfect agreement between graders for the measurement of GA area for all 3 modalities (intraclass correlation coefficient: 0.996 for manual Optos Advance, 0.996 for manual Heidelberg HEYEX, and 0.995 for Heidelberg Region Finder). The measurement of GA area was strongly correlated between modalities, with Spearman correlation coefficients of 0.985 (P < 0.001) between manual Heidelberg and manual Optos, 0.991 (P < 0.001) for Region Finder versus manual Heidelberg, and 0.985 (P < 0.001) for Region Finder versus manual Optos. The absolute mean area differences between the Heidelberg manual versus Region Finder, manual Optos versus Region Finder, and manual Optos versus manual Heidelberg were 1.61 mm2 (P < 0.001), 0.90 mm2 (P < 0.006), and 0.71 mm2 (P<0.001), respectively. We observed excellent interreader agreement for measurement of GA using either 30-degree blue-light FAF or UWF green-light FAF, establishing the reliability of UWF imaging for macular GA assessment. Although the absolute measurements between devices were strongly correlated, they differed significantly, highlighting the importance of using the same device for a given patient for the duration of a study. Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Compound Acoustic Radiation Force Impulse Imaging of Bovine Eye by Using Phase-Inverted Ultrasound Transducer.

In general, it is difficult to visualize internal ocular structure and detect a lesion such as a cataract or glaucoma using the current ultrasound brightness-mode (B-mode) imaging. This is because the internal structure of the eye is rich in moisture, resulting in a lack of contrast between tissues in the B-mode image, and the penetration depth is low due to the attenuation of the ultrasound wave. In this study, the entire internal ocular structure of a bovine eye was visualized in an ex vivo environment using the compound acoustic radiation force impulse (CARFI) imaging scheme based on the phase-inverted ultrasound transducer (PIUT). In the proposed method, the aperture of the PIUT is divided into four sections, and the PIUT is driven by the out-of-phase input signal capable of generating split-focusing at the same time. Subsequently, the compound imaging technique was employed to increase signal-to-noise ratio (SNR) and to reduce displacement error. The experimental results demonstrated that the proposed technique could provide an acoustic radiation force impulse (ARFI) image of the bovine eye with a broader depth-of-field (DOF) and about 80% increased SNR compared to the conventional ARFI image obtained using the in-phase input signal. Therefore, the proposed technique can be one of the useful techniques capable of providing the image of the entire ocular structure to diagnose various eye diseases.

Non-invasive coronary artery disease identification through the iris and bio-demographic health profile features using stacking learning

This study proposes a non-invasive method for predicting Coronary Artery Disease (CAD) using iris analysis, patient data, and Machine Learning (ML), primarily with iris images. It involved 281 participants, comprising 155 CAD patients and 126 non-patient controls, with eye images and biodemographic data collected at a Cardiology outpatient clinic. The study explored three scenarios: Scenario-I focused on biodemographic data, Scenario-II on iris features, and Scenario-III combined iris images and data. Iris processing included location determination, normalization, and heart region selection, with image enhancement via adaptive histogram equalization. Feature extraction through a 2-level wavelet transform generated 272 attributes, including statistical, Gray Level Co-occurrence Matrix, and Gray Level Run Length Matrix features for eight subcomponents. Correlation-based selection identified the best features, and classification employed ML techniques and incorporated stacking learning to enhance the results. Scenario-I achieved the highest accuracy at 83.57% among all evaluated algorithms. In Scenario-II, the proposed algorithm consistently outperformed others, achieving 94.88% accuracy and strong performance in other metrics, highlighting its effectiveness. In Scenario-III, the algorithm maintained superiority with 96.07% accuracy, specificity, recall, and area under the curve values. The proposed algorithm consistently outperforms other methods across scenarios, indicating its potential for CAD diagnosis, making it a promising choice for future CAD systems. The proposed algorithm presents a novel approach to the preliminary diagnosis of CAD, eliminating the necessity for electrocardiography, echocardiography, or effort tests. It also enables seamless integration into telemedicine systems, allowing for tele-diagnosis to conduct preliminary assessments before routine clinical practice.

Implementation of a High-Accuracy Neural Network-Based Pupil Detection System for Real-Time and Real-World Applications.

In this paper, the implementation of a new pupil detection system based on artificial intelligence techniques suitable for real-time and real-word applications is presented. The proposed AI-based pupil detection system uses a classifier implemented with slim-type neural networks, with its classes being defined according to the possible positions of the pupil within the eye image. In order to reduce the complexity of the neural network, a new parallel architecture is used in which two independent classifiers deliver the pupil center coordinates. The training, testing, and validation of the proposed system were performed using almost 40,000 eye images with a resolution of 320 × 240 pixels and coming from 20 different databases, with a high degree of generality. The experimental results show a detection rate of 96.29% at five pixels with a standard deviation of 3.38 pixels for all eye images from all databases and a processing speed of 100 frames/s. These results indicate both high accuracy and high processing speed, and they allow us to use the proposed solution for different real-time applications in variable and non-uniform lighting conditions, in fields such as assistive technology to communicate with neuromotor-disabled patients by using eye typing, in computer gaming, and in the automotive industry for increasing traffic safety by monitoring the driver's cognitive state.

How small changes to one eye’s retinal image can transform the perceived shape of a very familiar object

Vision can provide useful cues about the geometric properties of an object, like its size, distance, pose, and shape. But how the brain merges these properties into a complete sensory representation of a three-dimensional object is poorly understood. To address this gap, we investigated a visual illusion in which humans misperceive the shape of an object due to a small change in one eye's retinal image. We first show that this illusion affects percepts of a highly familiar object under completely natural viewing conditions. Specifically, people perceived their own rectangular mobile phone to have a trapezoidal shape. We then investigate the perceptual underpinnings of this illusion by asking people to report both the perceived shape and pose of controlled stimuli. Our results suggest that the shape illusion results from distorted cues to object pose. In addition to yielding insights into object perception, this work informs our understanding of how the brain combines information from multiple visual cues in natural settings. The shape illusion can occur when people wear everyday prescription spectacles; thus, these findings also provide insight into the cue combination challenges that some spectacle wearers experience on a regular basis.

An Investigation of a Biomimetic Optical System and an Evaluation Model for the Qualitative Analysis of Laser Interference Visual Levels.

To objectively quantify the level of visual interference induced by lasers, we developed a biomimetic optical system designed to emulate human vision. This system is based on an optical model of the eye and synthetic imaging principles, allowing it to generate biomimetic optical images that closely mimic human visual perception. Upon exposure to a 532 nm laser, biomimetic optical images were captured under various ambient lighting conditions. By employing a contrast threshold model for human visual target detection and grayscale hierarchy analysis, we devised an evaluation model to quantify the levels of laser-induced visual interference. The bionic images obtained from our experiments, in conjunction with the constructed model, enabled us to assess the degree of laser-induced visual interference. Our results indicate that this system can effectively substitute the human eye when testing laser imaging effects, with the generated bionic images achieving up to 90% concordance with human vision. The proposed evaluation model facilitates the quantitative analysis of laser-induced visual impairment. This apparatus and evaluation model hold significant promise for the precise quantification of laser-induced visual interference levels.

Comparison between Two Adaptive Optics Methods for Imaging of Individual Retinal Pigmented Epithelial Cells.

The Retinal Pigment Epithelium (RPE) plays a prominent role in diseases such as age-related macular degeneration, but imaging individual RPE cells is challenging due to their high absorption and low autofluorescence emission. The RPE lies beneath the highly reflective photoreceptor layer (PR) and contains absorptive pigments, preventing direct backscattered light detection when the PR layer is intact. Here, we used near-infrared autofluorescence adaptive optics scanning laser ophthalmoscopy (NIRAF AOSLO) and transscleral flood imaging (TFI) in the same healthy eyes to cross-validate these approaches. Both methods revealed a consistent RPE mosaic pattern and appeared to reflect a distribution of fluorophores consistent with findings from histological studies. Interestingly, even in apparently healthy RPE, we observed dynamic changes over months, suggesting ongoing cellular activity or alterations in fluorophore distribution. These findings emphasize the value of NIRAF AOSLO and TFI in understanding RPE morphology and dynamics.

Comparison of veterinarians and a deep learning tool in the diagnosis of equine ophthalmic diseases.

The aim was to compare ophthalmic diagnoses made by veterinarians to a deep learning (artificial intelligence) software tool which was developed to aid in the diagnosis of equine ophthalmic diseases. As equine ophthalmology is a very specialised field in equine medicine, the tool may be able to help in diagnosing equine ophthalmic emergencies such as uveitis. In silico tool development and assessment of diagnostic performance. A deep learning tool which was developed and trained for classification of equine ophthalmic diseases was tested with 40 photographs displaying various equine ophthalmic diseases. The same data set was shown to different groups of veterinarians (equine, small animal, mixed practice, other) using an opinion poll to compare the results and evaluate the performance of the programme. Convolutional Neural Networks (CNN) were trained on 2346 photographs of equine eyes, which were augmented to 9384 images. Two hundred and sixty-one separate unmodified images were used to evaluate the trained network. The trained deep learning tool was used on 40 photographs of equine eyes (10 healthy, 12 uveitis, 18 other diseases). An opinion poll was used to evaluate the diagnostic performance of 148 veterinarians in comparison to the software tool. The probability for the correct answer was 93% for the AI programme. Equine veterinarians answered correctly in 76%, whereas other veterinarians reached 67% probability for the correct diagnosis. Diagnosis was solely based on images of equine eyes without the possibility to evaluate the inner eye. The deep learning tool proved to be at least equivalent to veterinarians in assessing ophthalmic diseases in photographs. We therefore conclude that the software tool may be useful in detecting potential emergency cases. In this context, blindness in horses may be prevented as the horse can receive accurate treatment or can be sent to an equine hospital. Furthermore, the tool gives less experienced veterinarians the opportunity to differentiate between uveitis and other ocular anterior segment disease and to support them in their decision-making regarding treatment.

Framework for fish freshness detection and rotten fish removal in Bangladesh using mask R–CNN method with robotic arm and fisheye analysis

Fish-exporting countries must meet international standards and customer expectations to avoid trade disruption and economic repercussions. Ensuring the quality of fish products is therefore not just a matter of reputation but also of economic stability. Bangladesh, a major fish exporter, maintaining the quality of exported fish products is crucial. A single faulty product can have serious consequences, potentially causing harm. For this reason, we develop a deep learning-based approach capable of detecting and dividing rotten fish. We have used the Mask R–CNN method for our model. A device captures images of fish eyes and sends them to a computer system. The condition of a fish, whether fresh or rotten, can be determined by its eyes. A collection of 5000 image datasets is developed in this research work. The input images start matching the dataset through the Mask R–CNN and give us the result. Based on the result, if the fish is found fresh, it will proceed through the conveyor system; however, if it is identified as rotten, a robotic arm will separate it. To test its efficiency and reliability, we have tested it with 5000 images. And the test satisfied us with the results of 96.5% accuracy. With the assistance of our model, fish-exporting nations can efficiently distinguish between fresh and rotten fish, enhancing their ability to export more significant quantities of high-quality fish.

Gaze Estimation by Attention-Induced Hierarchical Variational Auto-Encoder.

Appearance-based gaze estimation has been widely studied recently with promising performance. The majority of appearance-based gaze estimation methods are developed under the deterministic frameworks. However, the deterministic gaze estimation methods suffer from large performance drop upon challenging eye images in low-resolution, darkness, partial occlusions, etc. To alleviate this problem, in this article, we alternatively reformulate the appearance-based gaze estimation problem under a generative framework. Specifically, we propose a variational inference model, that is, variational gaze estimation network (VGE-Net), to generate multiple gaze maps as complimentary candidates simultaneously supervised by the ground-truth gaze map. To achieve robust estimation, we adaptively fuse the gaze directions predicted on these candidate gaze maps by a regression network through a simple attention mechanism. Experiments on three benchmarks, that is, MPIIGaze, EYEDIAP, and Columbia, demonstrate that our VGE-Net outperforms state-of-the-art gaze estimation methods, especially on challenging cases. Comprehensive ablation studies also validate the effectiveness of our contributions. The code will be publicly released.

FEATURES OF THE MODIFICATION OF THE INCEPTIONRESNETV2 ARCHITECTURE AND THE CREATION OF A DIAGNOSTIC SYSTEM FOR DETERMINING THE DEGREE OF DAMAGE TO RETINAL VESSELS

Diabetic retinopathy is a retinal disease caused by diabetes. The progression of this disease can lead to blindness. Every year, the number of patients with this disease increases. Diabetic retinal damage can be slowed if it is diagnosed early. The article describes the features of the creation of a neural network model and the development of a system with high accuracy rates for the recognition of diabetic retinopathy. The advantages of the InceptionResNetv2 convolutional neural network architecture are considered. This network uses residual connections that help facilitate the learning process. InceptionResNetv2 uses different methods to reduce the dimensionality of the feature map, making it more economical in terms of memory and computation. This model has a number of advantages compared to other networks. InceptionResNetv2 can be applied to blood vessel segmentation in eye images with different resolutions. In the study, modification of InceptionResNetv2 was carried out. The use of additional MaxPooling and Dense layers improved the speed and accuracy of the InceptionResNetv2 convolutional neural network. The Dropout layer is effectively used to prevent overtraining. The system for determining the degree of retinal damage of diabetic origin is implemented in the Python programming language. Model layers are built using the Keras library. Images from the set of EyePacs were processed by methods of cropping the black frames with a Gaussian blur filter and minimizing the effect of changing the position of the images. During the research, it was found that 21 epochs are needed to achieve maximum accuracy. The program calculates the probability of an image belonging to a certain class with high accuracy. The recognition accuracy rate for class 1 was 98.6%, for class 2 - 98.5%, for class 3 - 98.3%, for class 4 - 98.15%, for class 5 - 98.1%.

Progress and challenges of in vivo flow cytometry and its applications in circulating cells of eyes.

Circulating inflammatory cells in eyes have emerged as early indicators of numerous major diseases, yet the monitoring of these cells remains an underdeveloped field. In vivo flow cytometry (IVFC), a noninvasive technique, offers the promise of real-time, dynamic quantification of circulating cells. However, IVFC has not seen extensive applications in the detection of circulating cells in eyes, possibly due to the eye's unique physiological structure and fundus imaging limitations. This study reviews the current research progress in retinal flow cytometry and other fundus examination techniques, such as adaptive optics, ultra-widefield retinal imaging, multispectral imaging, and optical coherence tomography, to propose novel ideas for circulating cell monitoring.

Exploiting Eye Colors for Better Iris Segmentation in Visible Wavelength Environments

Iris segmentation is a crucial step in iris recognition systems. Iris segmentation in visible wavelength and unconstrained environments is more challenging than ‎segmenting iris images in ideal environments. This paper proposes a new iris segmentation method that exploits the color of human eyes to segment the iris region more accurately. While most of the current iris segmentation methods ignore the color of the iris or deal with ‎grayscale eye images directly, the proposed method takes benefits from iris color to simplify the iris segmentation step. In the first step, we estimate the expected iris center using Haar-like features. The iris color is detected and accordingly, a color-convenient segmentation algorithm is applied to find the iris region. Dealing separately with each iris color set significantly decreases the false segmentation errors and enhances the performance of the iris recognition system. The results of testing the proposed algorithm on the UBIRIS database demonstrate the robustness of our algorithm against different noise factors and non-ideal conditions.