Glaucoma is one of the most common causes of blindness in the world. Screening glaucoma from retinal fundus images based on deep learning is a common method at present. In the diagnosis of glaucoma based on deep learning, the blood vessels within the optic disc interfere with the diagnosis, and there is also some pathological information outside the optic disc in fundus images. Therefore, integrating the original fundus image with the vessel-removed optic disc image can improve diagnostic efficiency. In this paper, we propose a novel multi-step framework named MSGC-CNN that can better diagnose glaucoma. In the framework, (1) we combine glaucoma pathological knowledge with deep learning model, fuse the features of original fundus image and optic disc region in which the interference of blood vessel is specifically removed by U-Net, and make glaucoma diagnosis based on the fused features. (2) Aiming at the characteristics of glaucoma fundus images, such as small amount of data, high resolution, and rich feature information, we design a new feature extraction network RA-ResNet and combined it with transfer learning. In order to verify our method, we conduct binary classification experiments on three public datasets, Drishti-GS, RIM-ONE-R3, and ACRIMA, with accuracy of 92.01%, 93.75%, and 97.87%. The results demonstrate a significant improvement over earlier results.

Cataracts and glaucoma account for a high percentage of vision loss and blindness worldwide. Small extracellular vesicles (sEVs) are released into different body fluids, including the eye\'s aqueous humor. Information about their proteome content and characterization in ocular pathologies is not yet well established. In this study, aqueous humor sEVs from healthy individuals, cataracts, and glaucoma patients were studied, and their specific protein profiles were characterized. Moreover, the potential of identified proteins as diagnostic glaucoma biomarkers was evaluated. The protein content of sEVs from patients\' aqueous humor with cataracts and glaucoma compared to healthy individuals was analyzed by quantitative proteomics. Validation was performed by western blot (WB) and ELISA. A total of 828 peptides and 192 proteins were identified and quantified. After data analysis with the R program, 8 significantly dysregulated proteins from aqueous humor sEVs in cataracts and 16 in glaucoma showed an expression ratio ≥ 1.5. By WB and ELISA using directly aqueous humor samples, the dysregulation of 9 proteins was mostly confirmed. Importantly, GAS6 and SPP1 showed high diagnostic ability of glaucoma, which in combination allowed for discriminating glaucoma patients from control individuals with an area under the curve of 76.1% and a sensitivity of 65.6% and a specificity of 87.7%.

Glaucoma, a leading cause of blindness, damages the optic nerve, making early diagnosis challenging due to no initial symptoms. Fundus eye images taken with a non-mydriatic retinograph help diagnose glaucoma by revealing structural changes, including the optic disc and cup. This research aims to thoroughly analyze saliency maps in interpreting convolutional neural network decisions for diagnosing glaucoma from fundus images. These maps highlight the most influential image regions guiding the network\'s decisions. Various network architectures were trained and tested on 739 optic nerve head images, with nine saliency methods used. Some other popular datasets were also used for further validation. The results reveal disparities among saliency maps, with some consensus between the folds corresponding to the same architecture. Concerning the significance of optic disc sectors, there is generally a lack of agreement with standard medical criteria. The background, nasal, and temporal sectors emerge as particularly influential for neural network decisions, showing a likelihood of being the most relevant ranging from 14.55% to 28.16% on average across all evaluated datasets. We can conclude that saliency maps are usually difficult to interpret and even the areas indicated as the most relevant can be very unintuitive. Therefore, its usefulness as an explanatory tool may be compromised, at least in problems such as the one addressed in this study, where the features defining the model prediction are generally not consistently reflected in relevant regions of the saliency maps, and they even cannot always be related to those used as medical standards.

Metabolome analysis has gained widespread application in disease diagnosis owing to its ability to provide comprehensive information, including disease phenotypes. In this study, we utilized 3D superstructures fabricated through evaporation-induced microprinting to analyze the metabolome for glaucoma diagnosis. 3D superstructures offer the following advantages: high hotspot density per unit volume of the structure extending from two to three dimensions, excellent signal repeatability due to the reproducibility and defect tolerance of 3D printing, and high thermal stability due to the PVP-enclosed capsule form. Leveraging the superior optical properties of the 3D superstructure, we aimed to classify patients with glaucoma. The signal obtained from the 3D superstructure was employed in a Deep Neural Network (DNN) classification model to accurately classify glaucoma patients. The sensitivity and specificity of the model were determined as 92.05% and 93.51%, respectively. Additionally, the fabrication of 3D superstructures can be performed at any stage, significantly reducing measurement time. Furthermore, their thermal stability allows for the analysis of smaller samples. One notable advantage of 3D superstructures is their versatility in accommodating different target materials. Consequently, they can be utilized for a wide range of metabolic analyses and disease diagnoses.

OBJECTIVE: The aim is to determine the magnitude of glaucomatous damage in the asymptomatic subjects identified with primary glaucoma for the first time and thus to evaluate the significance and efficacy of screening measures for glaucoma.
METHODS: An observational retrospective cohort study of 100 asymptomatic patients of age more than 40 years, diagnosed with and under treatment for primary glaucoma was performed. Patients were categorized into having early, moderate, and severe glaucoma, according to standard automated perimetry (SAP) mean deviation (MD) in the worse eye (<-6, -6 to -12 and >-12 dB, respectively). Risk factors were correlated with the severity of glaucoma at presentation and statistically analyzed.
RESULTS: About 32%, 33%, and 35% of patients were found to have early, moderate, and severe stages of glaucoma with average MD of -3.51 ±1.53, -8.65 ±1.64, -17.15 ± 5.13 on SAP, respectively. The association of risk factors such as age (P = 0.006) and glaucoma awareness (P = 0.044) with the severity of glaucoma was statistically significant. There was no direct statistical correlation found between gender, history of diabetes mellitus, family history of glaucoma, intraocular pressure, central corneal thickness, the angle width, and the severity of glaucoma in our study.
CONCLUSIONS: Majority of cases with primary glaucoma show no symptoms until advanced irreversible stages. Early screening and proper treatment are the only ways to halt its progression. In spite of available facilities, 68% of patients in our study were found to have moderate-to-severe stages of glaucoma. This indicates that our screening measures should reach the masses at the primary level, with a focus on awareness programs.

暂无摘要。

UNASSIGNED: To evaluate the efficacy of opportunistic case finding in glaucoma detection and to determine factors associated with failure of glaucoma detection by eye health providers.
UNASSIGNED: This study was conducted on 154 new definite primary open-angle glaucoma (POAG) patients presenting to our glaucoma clinic. A questionnaire was prepared to determine if these subjects had sought eye care up to 12 months before presentation. The type of eye care provider and the principal reason for the visit were probed. The primary outcome measure was the frequency of a correct glaucoma diagnosis in their index visit. The secondary outcomes were factors associated with missed POAG diagnosis.
UNASSIGNED: The great majority of study subjects (132 cases, 85.7%) had sought at least one ocular examination within 1 year before presentation. Among these patients, 73 cases (55.3%) had remained undiagnosed after the examination. Among the probed variables, age, gender, visual acuity, visual field defects, intraocular pressure, cup/disc ratio, nerve fiber layer thickness of the worse eye at presentation, and family history of glaucoma were comparable between correctly diagnosed and missed POAGs. The only factors significantly associated with missed POAG diagnosis were lack of significant refractive errors and visiting an optometrist rather than an ophthalmologist.
UNASSIGNED: The efficacy of opportunistic case finding for POAG seems to be less than ideal in our settings. Lack of a significant refractive error and visiting an optometrist rather than an ophthalmologist were associated with a missed diagnosis of POAG. These observations reflect the need to adopt policies to improve glaucoma screening by eye care providers.

Recent developments in the use of artificial intelligence in the diagnosis and monitoring of glaucoma are discussed. To set the context and fix terminology, a brief historic overview of artificial intelligence is provided, along with some fundamentals of statistical modeling. Next, recent applications of artificial intelligence techniques in glaucoma diagnosis and the monitoring of glaucoma progression are reviewed, including the classification of visual field images and the detection of glaucomatous change in retinal nerve fiber layer thickness. Current challenges in the direct application of artificial intelligence to further our understating of this disease are also outlined. The article also discusses how the combined use of mathematical modeling and artificial intelligence may help to address these challenges, along with stronger communication between data scientists and clinicians.

Glaucoma has become a major cause of vision loss. Early-stage diagnosis of glaucoma is critical for treatment planning to avoid irreversible vision damage. Meanwhile, interpreting the rapidly accumulated medical data from ophthalmic exams is cumbersome and resource-intensive. Therefore, automated methods are highly desired to assist ophthalmologists in achieving fast and accurate glaucoma diagnosis. Deep learning has achieved great successes in diagnosing glaucoma by analyzing data from different kinds of tests, such as peripapillary optical coherence tomography (OCT) and visual field (VF) testing. Nevertheless, applying these developed models to clinical practice is still challenging because of various limiting factors. OCT models present worse glaucoma diagnosis performances compared to those achieved by OCT&VF based models, whereas VF is time-consuming and highly variable, which can restrict the wide employment of OCT&VF models. To this end, we develop a novel deep learning framework that leverages the OCT&VF model to enhance the performance of the OCT model. To transfer the complementary knowledge from the structural and functional assessments to the OCT model, a cross-modal knowledge transfer method is designed by integrating a designed distillation loss and a proposed asynchronous feature regularization (AFR) module. We demonstrate the effectiveness of the proposed method for glaucoma diagnosis by utilizing a public OCT&VF dataset and evaluating it on an external OCT dataset. Our final model with only OCT inputs achieves the accuracy of 87.4% (3.1% absolute improvement) and AUC of 92.3%, which are on par with the OCT&VF joint model. Moreover, results on the external dataset sufficiently indicate the effectiveness and generalization capability of our model.

To compare the Swedish Interactive Thresholding Algorithm (SITA) Standard (SS) and SITA Faster (SFR) strategies in normal individuals undergoing standard automated perimetry (SAP) for the first time.
Randomized, comparative, observational case series.
Seventy-four perimetry-naive healthy individuals.
All individuals underwent SAP 24-2 testing with the Humphrey Field Analyzer III (model 850 Zeiss) using the SS and SFR strategies. One eye of each individual was tested. Test order between strategies was randomized, and an interval of 15 minutes was allowed between the tests.
The following variables were compared: test time, foveal threshold, false-positive errors, number of unreliable tests, mean deviation (MD), visual field index (VFI), pattern standard deviation (PSD), glaucoma hemifield test (GHT), and number of depressed points deviating at P < 5%, P < 2%, P < 1%, and P < 0.5% on the total and pattern deviation probability maps. Specificity of the SS and SFR strategies were compared using Anderson\'s criteria for abnormal visual fields.
The SFR tests were 60.4% shorter in time compared with SS (P < 0.001) and were associated with a significantly lower PSD (1.75 ± 0.80 decibel [dB] vs. 2.15 ± 1.25 dB; P = 0.016). There were no significant differences regarding the MD, VFI, foveal threshold, GHT, and number of points depressed at P < 5%, P < 2%, P < 1%, and P < 0.5% on the total deviation and pattern deviation probability maps between SS and SFR. When all exams were analyzed and any of Anderson\'s criteria was applied, the specificity was 68% with SFR and 61% with SS (P = 0.250). The specificities observed with SFR and SS when only the first or second exams were analyzed were also similar (63% vs. 64% and 72% vs. 58%, respectively, P > 0.05).
The SS and SFR were associated with similar specificities in perimetry-naive individuals. The SFR did not increase the number of depressed points in the total and pattern deviation probability maps. Ophthalmologists should be aware that both strategies are associated with disturbingly high false-positive rates in perimetry-naive individuals.
Proprietary or commercial disclosure may be found after the references.