Uveoscleral Outflow Routes after MicroPulse Laser Therapy for Refractory Glaucoma: An Optical Coherence Tomography Study of the Sclera.

To analyze in vivo scleral changes induced by MicroPulse transscleral laser therapy (MP-TLT) in refractory glaucoma using anterior segment-optical coherence tomography (AS-OCT). Forty-two candidate patients for MP-TLT were consecutively enrolled and underwent AS-OCT at baseline and after six months. MP-TLT success was defined as an intraocular pressure (IOP) reduction by one-third. The main outcome measures were the mean superior (S-), inferior (I-), and total (T-) intra-scleral hypo-reflective space area (MISHA: mm2) and scleral reflectivity (S-SR, I-SR, T-SR; arbitrary scale) as in vivo biomarkers of uveoscleral aqueous humor (AH) outflow. The IOP was the secondary outcome. The relations between the baseline-to-six months differences (D) of DS-MISHA, DI-MISHA, and DT-MISHA and DS-SR, DI-SR, DT-SR, and DIOP, were investigated. At 6 months, the median IOP reduction was 21% in the failures and 38% in the successes. The baseline S-MISHA, I-MISHA, and T-MISHA did not differ between the groups, while S-SR and T-SR were higher in the successes (p < 0.05). At six months, successful and failed MP-TLTs showed a 50% increase in S-MISHA (p < 0.001; p = 0.037), whereas I-SR and T-SR reduced only in the successes (p = 0.002; p = 0.001). When comparing DS-MISHA, DI-MISHA, and DT-MISHA and DS-SR, DI-SR, and DT-SR, there were no significant differences between the groups. In the successful procedures, DIOP was positively correlated with DT-MISHA and DI-MISHA (ρ = 0.438 and ρ = 0.490; p < 0.05). MP-TLT produced potentially advantageous modifications of the sclera in refractory glaucoma. Given the partial correlation between these modifications and post-treatment IOP reduction, our study confirmed that the activation of the uveoscleral AH outflow route could significantly contribute to the IOP lowering after MP-TLT.

Bleb characteristics assessed by ultrasound biomicroscopy correlate with intraocular pressure after Ahmed glaucoma valve surgery in dogs.

OBJECTIVE: To assess the characteristics of blebs formed after Ahmed glaucoma valve (AGV) surgery in dogs using ultrasound biomicroscopy (UBM) and to analyze their correlation with postoperative intraocular pressure (IOP). ANIMALS: 16 eyes (13 dogs) were diagnosed with primary angle-closure glaucoma and were followed up after AGV surgery from June 2021 to September 2023. METHODS: In this prospective study, UBM examinations were performed to assess bleb characteristics, including bleb wall thickness and reflectivity. IOP at the time of UBM imaging and the duration from AGV surgery to UBM imaging were recorded. Histological examination of an enucleated eye removed due to uncontrolled IOP leading to blindness was also conducted. RESULTS: A significant correlation was observed between IOP and relative reflectivity (Pearson r = 0.60; P = .01), and a negative correlation was observed between bleb wall thickness and relative reflectivity (Pearson r = -0.72; P = .002). No significant correlation was observed between the duration from AGV surgery to UBM imaging and either bleb wall thickness or relative reflectivity, respectively. Histological examination of the enucleated eye revealed collagen-rich fibrous encapsulation of the bleb wall, including myofibroblasts that exhibited positive α-smooth muscle actin immunostaining. CLINICAL RELEVANCE: In dogs that underwent AGV surgery, less dense, thick-walled blebs on UBM tended to maintain IOP within the normal range. However, denser, thinner-walled blebs showed IOP levels above the normal range despite the use of antiglaucoma medications. UBM is a useful tool for evaluating bleb characteristics and their influence on IOP regulation after AGV surgery in dogs.

Artificial intelligence in glaucoma: opportunities, challenges, and future directions.

Artificial intelligence (AI) has shown excellent diagnostic performance in detecting various complex problems related to many areas of healthcare including ophthalmology. AI diagnostic systems developed from fundus images have become state-of-the-art tools in diagnosing retinal conditions and glaucoma as well as other ocular diseases. However, designing and implementing AI models using large imaging data is challenging. In this study, we review different machine learning (ML) and deep learning (DL) techniques applied to multiple modalities of retinal data, such as fundus images and visual fields for glaucoma detection, progression assessment, staging and so on. We summarize findings and provide several taxonomies to help the reader understand the evolution of conventional and emerging AI models in glaucoma. We discuss opportunities and challenges facing AI application in glaucoma and highlight some key themes from the existing literature that may help to explore future studies. Our goal in this systematic review is to help readers and researchers to understand critical aspects of AI related to glaucoma as well as determine the necessary steps and requirements for the successful development of AI models in glaucoma.

A fast and fully automated system for glaucoma detection using color fundus photographs.

This paper presents a low computationally intensive and memory efficient convolutional neural network (CNN)-based fully automated system for detection of glaucoma, a leading cause of irreversible blindness worldwide. Using color fundus photographs, the system detects glaucoma in two steps. In the first step, the optic disc region is determined relying upon You Only Look Once (YOLO) CNN architecture. In the second step classification of 'glaucomatous' and 'non-glaucomatous' is performed using MobileNet architecture. A simplified version of the original YOLO net, specific to the context, is also proposed. Extensive experiments are conducted using seven state-of-the-art CNNs with varying computational intensity, namely, MobileNetV2, MobileNetV3, Custom ResNet, InceptionV3, ResNet50, 18-Layer CNN and InceptionResNetV2. A total of 6671 fundus images collected from seven publicly available glaucoma datasets are used for the experiment. The system achieves an accuracy and F1 score of 97.4% and 97.3%, with sensitivity, specificity, and AUC of respectively 97.5%, 97.2%, 99.3%. These findings are comparable with the best reported methods in the literature. With comparable or better performance, the proposed system produces significantly faster decisions and drastically minimizes the resource requirement. For example, the proposed system requires 12 times less memory in comparison to ResNes50, and produces 2 times faster decisions. With significantly less memory efficient and faster processing, the proposed system has the capability to be directly embedded into resource limited devices such as portable fundus cameras.

C2FTFNet: Coarse-to-fine transformer network for joint optic disc and cup segmentation.

Glaucoma is a leading cause of worldwide blindness and visual impairment, making early screening and diagnosis is crucial to prevent vision loss. Cup-to-Disk Ratio (CDR) evaluation serves as a widely applied approach for effective glaucoma screening. At present, deep learning methods have exhibited outstanding performance in optic disk (OD) and optic cup (OC) segmentation and maturely deployed in CAD system. However, owning to the complexity of clinical data, these techniques could be constrained. Therefore, an original Coarse-to-Fine Transformer Network (C2FTFNet) is designed to segment OD and OC jointly , which is composed of two stages. In the coarse stage, to eliminate the effects of irrelevant organization on the segmented OC and OD regions, we employ U-Net and Circular Hough Transform (CHT) to segment the Region of Interest (ROI) of OD. Meanwhile, a TransUnet3+ model is designed in the fine segmentation stage to extract the OC and OD regions more accurately from ROI. In this model, to alleviate the limitation of the receptive field caused by traditional convolutional methods, a Transformer module is introduced into the backbone to capture long-distance dependent features for retaining more global information. Then, a Multi-Scale Dense Skip Connection (MSDC) module is proposed to fuse the low-level and high-level features from different layers for reducing the semantic gap among different level features. Comprehensive experiments conducted on DRIONS-DB, Drishti-GS, and REFUGE datasets validate the superior effectiveness of the proposed C2FTFNet compared to existing state-of-the-art approaches.

Bleb analysis using anterior segment optical coherence tomography after trabeculectomy with amniotic membrane transplantation.

INTRODUCTION: Little has been known about the intrableb structures associated with bleb function after trabeculectomy with amniotic membrane transplantation (AMT). The aim of this study is to analyze the characteristics of intrableb structures using anterior segment optical coherence tomography (AS-OCT) after trabeculectomy with AMT. METHODS: A total of 68 eyes of 68 patients with primary open-angle glaucoma who underwent trabeculectomy with AMT were included. Surgical success was defined as intraocular pressure (IOP) ≤ 18 mmHg and IOP reduction of ≥ 20% without medication on AS-OCT examination. Intrableb parameters, including bleb height, bleb wall thickness, striping layer thickness, bleb wall reflectivity, fluid-filled space score, fluid-filled space height, and microcyst formation were evaluated using AS-OCT. Logistic regression analysis was performed to determine factors associated with IOP control. RESULTS: Of the 68 eyes, 56 eyes were assigned to the success group and 12 eyes to the failure group. In the success group, bleb height (P = 0.009), bleb wall thickness (P = 0.001), striping layer thickness (P = 0.001), fluid-filled space score (P = 0.001), and frequency of microcyst formation (P = 0.001) were greater than those in the failure group. Bleb wall reflectivity was higher in the failure group than in the success group (P < 0.001). In the univariate logistic regression analysis, previous cataract surgery was significantly associated with surgical failure (odds ratio = 5.769, P = 0.032). CONCLUSION: A posteriorly extending fluid-filled space, tall bleb with low reflectivity, and thick striping layer were characteristics of successful filtering blebs after trabeculectomy with AMT.

Joint optic disc and cup segmentation based on elliptical-like morphological feature and spatial geometry constraint.

Glaucoma is a chronic degenerative disease that is the second leading cause of irreversible blindness worldwide. For a precise and automatic screening of glaucoma, detecting the optic disc and cup precisely is significant. In this paper, combining the elliptical-like morphological features of the disc and cup, we reformulate the segmentation task from a perspective of ellipse detection to explicitly segment and directly get the glaucoma screening indicator. We detect the minimum bounding boxes of ellipses firstly, and then learn the ellipse parameters of these regions to achieve optic disc and cup segmentation. Considering the spatial geometry prior knowledge that the cup should be within the disc region, Paired-Box RPN is introduced to simultaneously detect the disc and cup coupled. In addition, boundary attention module is introduced to use edges of the disc and cup as an important guide for context aggregation to improve the accuracy. Comprehensive experiments clearly show that our method outperforms the state-of-the-art methods for optic disc and cup segmentation. Simultaneously, the proposed method also obtains the good glaucoma screening performance with calculated vCDR value. Joint optic disc and cup segmentation, which utilizes the elliptical-like morphological features and spatial geometry constraint, could improve the performance of optic disc and cup segmentation.

Clinical applications of anterior segment swept-source optical coherence tomography: A systematic review.

Optical coherence tomography (OCT) is a non-invasive method that provides the opportunity to examine tissues by taking cross-sectional images. OCT is increasingly being used to evaluate anterior segment (AS) pathologies. Swept-source (SS) OCT allows greater penetration and achieves better visualization of the internal configuration of AS tissues due to the longer wavelength employed and high scan speeds. We reviewed the utilization of AS SS-OCT in various conditions including glaucoma, ocular surface pathologies, iris tumors, refractive surgery, cataract surgery, and scleral diseases. A systematic literature search was carried out on PubMed, Scopus, and Web of Science databases between January 1, 2008, and September 1, 2022 using the following keywords: AS SS-OCT; dry eye and SS-OCT; ocular surface and SS-OCT; cornea and SS-OCT; dystrophy and SS-OCT; glaucoma and SS-OCT; ocular surface tumors and SS-OCT; conjunctival tumors and SS-OCT; refractive surgery and SS-OCT; cataract and SS-OCT; biometry and SS-OCT; sclera and SS-OCT; iris and SS-OCT; ciliary body and SS-OCT; artificial intelligence and SS-OCT. A total of 221 studies were included in this review. Review of the existing literature shows that SS-OCT offers several advantages in the diagnosis of AS diseases. Exclusive features of SS-OCT including rapid scanning, deeper tissue penetration, and better image quality help improve our understanding of various AS pathologies.

Ganglion cell complex thickness with spectral domain optical coherence tomography and correlations in a normative pediatric South Asian cohort.

Spectral domain optical coherence tomography (SD-OCT) was used to assess the perifoveal Ganglion cell complex (GCC) thickness in healthy Pakistani children. 174 eyes of 87 healthy children (3 to 16 years) were included after a thorough eye examination. Significant refractive error was excluded (>± 5 DS), glaucoma, ocular pathology, intraocular surgery, trauma, and systemic ailments. The perifoveal GCC analysis was done by Optopol Revo 80® high resolution SD-OCT. The average age was 11.1 ± 3.12 years. The average GCC thickness globally was 120.01 ± 76.74 µm, superiorly; 117.51 ± 8.77 µm, superonasally; 120.47 ± 8.78 µm, superotemporally, 103.83 ± 8.25 µm, inferiorly; 117.41 ± 8.71 µm, inferonasally 120.57 ± 8.88 µm, and inferotemporally; 107.15 ± 7.69 µm. The average GCC thickness was positively correlated with axial length (p = .02) and the GCC thickness in all sectors was significantly higher amongst males as compared to females (p < .05). Our study provides normative data for the macular GCC thickness in healthy children of our population. GCC thickness estimation is an important diagnostic parameter in glaucoma and optic neuropathies. Establishing a normative database for our pediatric population would help immensely in the diagnosis and monitoring of optic nerve disorders. RESEARCH HIGHLIGHTS: Spectral domain optical coherence tomography guided measurement of the ganglion cell complex thickness of the Pakistani pediatric population has not been done before. Our study provides normative data for the macular ganglion cell complex in healthy children of our population and studies its correlations with age, gender, intraocular pressure, axial length, refractive status, and corneal thickness. This will guide us in optic nerve and neurological disease diagnosis and monitoring.

Changes in the deep vasculature assessed using anterior segment OCT angiography following trabecular meshwork targeted minimally invasive glaucoma surgery.

The effect of trabecular meshwork (TM)-targeted minimally invasive glaucoma surgery (MIGS) on the vasculature assessed using anterior segment (AS)-optical coherence tomography angiography (OCTA) has not been established. In this prospective, longitudinal study, we investigated changes in the deep vasculature following TM-targeted MIGS using AS-OCTA for open-angle glaucoma in 31 patients. AS-OCTA images of the sclera and conjunctiva at the nasal corneal limbus were acquired preoperatively and 3 months postoperatively, and the vessel densities (VDs) of the superficial (conjunctival) and deep (intrascleral) layers were calculated. The VDs before and after MIGS were compared, and the factors associated with the change in VD following MIGS were analyzed. The mean deep VD decreased from 11.98 ± 6.80% at baseline to 10.42 ± 5.02% postoperatively (P = 0.044), but superficial VD did not change (P = 0.73). The multivariate stepwise regression analysis revealed that deep VD reduction was directly associated with IOP reduction (P < 0.001) and preoperative IOP (P = 0.007) and inversely associated with preoperative deep VD (P < 0.001). The deep VD reduction following MIGS was significant in the successful group (21 eyes) (P = 0.032) but not in the unsuccessful group (10 eyes) (P = 0.49). The deep VDs assessed using AS-OCTA decreased following TM-targeted MIGS, especially in the eyes with good surgical outcomes.

Glaucoma screening using an attention-guided stereo ensemble network.

Glaucoma is a chronic eye disease, which causes gradual vision loss and eventually blindness. Accurate glaucoma screening at early stage is critical to mitigate its aggravation. Extracting high-quality features are critical in training of classification models. In this paper, we propose a deep ensemble network with attention mechanism that detects glaucoma using optic nerve head stereo images. The network consists of two main sub-components, a deep Convolutional Neural Network that obtains global information and an Attention-Guided Network that localizes optic disc while maintaining beneficial information from other image regions. Both images in a stereo pair are fed into these sub-components, the outputs are fused together to generate the final prediction result. Abundant image features from different views and regions are being extracted, providing compensation when one of the stereo images is of poor quality. The attention-based localization method is trained in a weakly-supervised manner and only image-level annotation is required, which avoids expensive segmentation labelling. Results from real patient images show that our approach increases recall (sensitivity) from the state-of-the-art 88.89% to 95.48%, while maintaining precision and performance stability. The marked reduction in false-negative rate can significantly enhance the chance of successful early diagnosis of glaucoma.

New views on three-dimensional imaging technologies for glaucoma: an overview.

PURPOSE OF REVIEW: To summarize the literature on three-dimensional (3D) technological advances in ophthalmology, the quantitative methods associated with this, and their improved ability to help detect glaucoma disease progression. RECENT FINDINGS: Improvements in measuring glaucomatous structural changes are the result of dual innovations in optical coherence tomography (OCT) imaging technology and in associated quantitative software. SUMMARY: Compared with two-dimensional (2D) OCT parameters, newer 3D parameters provide more data and fewer artifacts.

Identifying and understanding optical coherence tomography artifacts that may be confused with glaucoma / Identificando e compreendendo os artefatos de tomografia de coerência óptica que podem ser confundidos com o glaucoma

ABSTRACT Optical coherence tomography is often used for detection of glaucoma as well as to monitor progression. This paper reviews the most common types of artifacts on the optical coherence tomography report that may be confused with glaucomatous damage. We mainly focus on anatomy-related artifacts in which the retinal layer segmentation and thickness measurements are correct. In such cases, the probability maps (also known as deviation maps) show abnormal (red and yellow) regions, which may mislead the clinician to assume disease is present. This is due to the anatomic variability of the individual, and the normative database must be taken into account.

RESUMO A tomografia de coerência óptica é frequentemente usada para detectar glaucoma, bem como para monitorar a progressão. Este artigo analisa os tipos mais comuns de artefatos no relatório de tomografia de coerência óptica que podem ser confundidos com danos glaucomatosos. Nós nos concentramos principalmente nos artefatos relacionados à anatomia em que a segmentação da camada da retina e as medidas de espessura estão corretas. Nesses casos, os mapas de probabilidade (também conhecidos como mapas de desvio) mostram regiões anormais (vermelho e amarelo), o que pode induzir o clínico em erro ao supor que a doença está presente. Isto se deve à variabilidade anatômica do indivíduo, e o banco de dados normativo deve ser levado em conta.

Longitudinal changes in complete avascular area assessed using anterior segmental optical coherence tomography angiography in filtering trabeculectomy bleb.

Optical coherence tomography angiography (OCTA) is a new technique for non-invasive imaging of blood vessels, allowing combined evaluation of both deep and surface vessels. The purpose of this study was to evaluate the post-trabeculectomy longitudinal changes in complete avascular area (CAA) of filtering blebs using anterior segment (AS-) OCTA and their association with surgical outcomes. This study included 57 eyes of 53 patients who had undergone trabeculectomy with mitomycin C. AS-OCTA images of filtering bleb were acquired at 3 and 6 months after trabeculectomy, and at 1 month in possible cases. CAAs, regions where complete blood flow was not depicted in AS-OCTA images, were evaluated for their presence, extent, and change over time. CAAs were detected in 37 eyes (65%) and 33 eyes (58%) at 3 and 6 months postoperatively, respectively. The extent of CAAs reduced over time after surgery in most cases. No parameters related to CAAs were significantly associated with surgical success (i.e., intraocular pressure (IOP) ≤ 12 mmHg and IOP reduction > 20% without medication). In conclusion, although it is difficult to predict surgical success by CAA itself, AS-OCTA may be useful for the objective evaluation of the vascularity of filtering blebs.

NENet: Nested EfficientNet and adversarial learning for joint optic disc and cup segmentation.

Glaucoma is an ocular disease threatening irreversible vision loss. Primary screening of Glaucoma involves computation of optic cup (OC) to optic disc (OD) ratio that is widely accepted metric. Recent deep learning frameworks for OD and OC segmentation have shown promising results and ways to attain remarkable performance. In this paper, we present a novel segmentation network, Nested EfficientNet (NENet) that consists of EfficientNetB4 as an encoder along with a nested network of pre-activated residual blocks, atrous spatial pyramid pooling (ASPP) block and attention gates (AGs). The combination of cross-entropy and dice coefficient (DC) loss is utilized to guide the network for accurate segmentation. Further, a modified patch-based discriminator is designed for use with the NENet to improve the local segmentation details. Three publicly available datasets, REFUGE, Drishti-GS, and RIM-ONE-r3 were utilized to evaluate the performances of the proposed network. In our experiments, NENet outperformed state-of-the-art methods for segmentation of OD and OC. Additionally, we show that NENet has excellent generalizability across camera types and image resolution. The obtained results suggest that the proposed technique has potential to be an important component for an automated Glaucoma screening system.

Establishment of appropriate glaucoma models using dexamethasone or TGFß2 treated three-dimension (3D) cultured human trabecular meshwork (HTM) cells.

To establish appropriate ex vivo models for a glaucomatous trabecular meshwork (TM), two-dimensional (2D) and three-dimensional (3D) cultures of human trabecular meshwork cells (HTM) were prepared in the presence of 250 nM dexamethasone (DEX) or 5 ng/mL TGFß2, and characterized by the following analyses; transendothelial electrical resistance (TEER) measurements, FITC dextran permeability, scanning electron microscopy and the expression of the extracellular matrix (ECM) including collagen (COL)1, 4 and 6, and fibronectin (FN), α-smooth muscle actin (α-SMA), tissue inhibitor of metalloproteinase (TIMP)1-4, and matrix metalloproteinase (MMP)2, 9 and 14. DEX and TGFß2 both caused a significant increase or decrease in the TEER values and FITC dextran permeability. During the 3D spheroid culture, DEX or TGFß2 induced a mild and significant down-sizing and an increase in stiffness, respectively. TGFß2 induced a significant up-regulation of COL1 and 4, FN, α-SMA, and MMP 2 and 14 (2D) or COL1 and 6, and TIMP2 and 3 (3D), and DEX induced a significant up-regulation of FN (3D) and TIMP4 (2D and 3D). The findings presented herein indicate that DEX or TGFß2 resulted in mild and severe down-sized and stiff 3D HTM spheroids, respectively, thus making them viable in vitro HTM models for steroid-induced and primary open angle glaucoma.

Prediction of trabecular meshwork-targeted micro-invasive glaucoma surgery outcomes using anterior segment OCT angiography.

We performed a prospective, longitudinal study to investigate the association between the preoperative intrascleral vasculature assessed using anterior segment (AS)-optical coherence tomography angiography (OCTA) and surgical outcomes of trabecular meshwork-targeted micro- or minimally invasive glaucoma surgery (MIGS). We included 37 patients with primary open-angle glaucoma. Preoperative AS-OCTA images of the sclero-conjunctiva of the nasal corneal limbus were acquired in the superficial (conjunctival) and deep (intrascleral) layers. The vessel densities (VDs) of each layer were measured separately in the entire area, limbal side, and fornix area. Surgical success was determined by postoperative intraocular pressure (IOP) and IOP reduction. Twenty-three and 14 eyes were classified as having successful and unsuccessful outcomes, respectively. The deep VDs of the entire area and fornix area were significantly lower in the successful group (P = 0.031 and P = 0.009). The success rate was significantly higher for eyes with a lower deep VD than for eyes with a higher deep VD. A greater IOP reduction was significantly associated with lower deep VD in the fornix area (P = 0.022) and higher preoperative IOP (P < 0.001). These results indicate that intrascleral vasculature assessed using preoperative AS-OCTA was negatively correlated with surgical success and IOP reduction resulting from trabecular meshwork-targeted MIGS. AS-OCTA images might help predict MIGS outcomes.

Use of computerized campimetry and/or optical coherence tomography for glaucoma diagnosis by non-glaucoma specialists / Uso da campimetria e/ou da tomografia de coerência óptica no diagnóstico do glaucoma por oftalmologistas generalistas

ABSTRACT Purpose: To compare the use of visual field and/or optical coherence tomography (OCT) combined with color retinography by non-glaucoma specialists for differentiating glaucoma from physiological cupping. Methods: Eighty patients with glaucoma or physiological cupping (40 of each) were randomized according to the examination used (GI: color retinography, GII: color retinography + visual field, GIII: color retinography + optical coherence tomography, GIV: color retinography + visual field + optical coherence tomography). Twenty non-specialist ophthalmologists diagnosed glaucoma from PowerPoint slide images, without direct patient examination. Results: Inter-examiner agreement was good for GII (ĸ: 0.63; 95%CI, 0.53-0.72), moderate for GIII (ĸ: 0.58; 95%CI, 0.48-0.68) and GIV (ĸ: 0.41; 95%CI, 0.31-0.51), and low for GI (ĸ: 0.30; 95%CI, 0.20-0.39) (p<0.001). Diagnostic accuracy was higher in GIII (15.8 ± 1.82) than GI (12.95 ± 1.46, p<0.001) and higher in GII (16.25 ± 2.02) than GI and GIV (14.10 ± 2.24) (both p<0.001). For glaucoma patients only, diagnostic accuracy in GII and GIII was superior to that in GI and GIV (both p<0.001). Sensitivity and specificity were 59% and 70.5% in GI; 86.5% and 76% in GII, 86.5% and 71.5% in GIII; and 68.5% and 72.5% in GIV, respectively. Accuracy was highest in GII (81.3% [95%CI, 77.1-84.8]), followed by GIII (79% [95%CI, 74.7-82.7]), GIV (70,5% [95%CI, 65.9-74.8]), and GI (64.8% [95%CI, 60.0-69.3]). Conclusions: Non-glaucoma specialists could not differentiate glaucoma from increased physiological cupping when using color retinography assessment alone. Diagnostic accuracy and inter-rater agreement improved significantly with the addition of visual field or optical coherence tomography. However, the use of both modalities did not improve sensitivity/specificity.(AU)

RESUMO Objetivos: Verificar a influência do campo visual e/ou tomografia de coerência óptica, quando analisados em associação à retinografia colorida, na diferenciação entre indivíduos com glaucoma daqueles com aumento fisiológico de escavação. Métodos: Oitenta pacientes com glaucoma ou aumento fisiológico de escavação (40 cada) foram randomizados de acordo com o exame testado (GI: retinografia colorida, GII: retinografia colorida + campo visual, GIII: retinografia colorida + tomografia de coerência óptica, GIV: retinografia colorida + campo visual + tomografia de coerência óptica). Vinte oftalmologistas não especialistas em glaucoma diagnosticaram glaucoma através de slides do PowerPoint, sem o exame direto do paciente. Resultados: A concordância interexaminador foi boa para o GII (ĸ: 0,63; 95%CI, 0,53-0,72), moderada para GIII (ĸ: 0,58; 95%CI, 0,48-0,68) e GIV (ĸ: 0,41; 95%CI, 0,31-0,51), e baixa para o GI (ĸ: 0,30; 95%CI, 0,20-0,39) (p<0,001). Acurácia diagnostica foi maior no GIII (15,8 ± 1,82) em comparação ao GI (12,95 ± 1,46, p<0,001) e o GII (16,25 ± 2,02) maior em comparação ao GI e GIV (14,10 ± 2,24) (para ambos, p<0,001). Para os pacientes com glaucoma, a acurácia diagnostica nos grupos GII e GIII foi superior do que em GI e GIV (ambos p<0,001). Sensibilidade e especificidade foram 59% e 70,5% no GI; 86,5% e 76% no GII, 86,5% e 71,5% no GIII; 68,5% e 72,5% no GIV, respectivamente. A acurácia foi maior no GII (81,3% [95%CI, 77,1-84,8]), seguido pelo GIII (79% [95%CI, 74,7-82,7]), GIV (70,5% [95%CI, 65,9-74,8]), e GI (64,8% [95%CI, 60,0-69,3]). Conclusões: A avaliação isolada da retinografia colorida por oftalmologistas não especialistas em glaucoma não pode diferenciar pacientes com glaucoma daqueles com aumento fisiológico de escavação. Houve aumento da acurácia diagnóstica e da concordância interobservador com o acréscimo do campo visual ou da tomografia de coerência óptica. Entretanto, o uso de ambas as modalidades não melhorou a sensibilidade/especificadade.(AU)

The Relationship Between Macula Retinal Ganglion Cell Density and Visual Function in the Nonhuman Primate.

Purpose: Loss of ganglion cell inner plexiform layer (GCIPL) and visual sensitivity in the macula region are known to occur at all stages of glaucoma. While both are dependent on the underlying retinal ganglion cells (RGCs), the relationship between structure and function is modest. We hypothesize that the imprecise relationship is due to a lack of direct correspondence between in vivo measures and RGC counts, as well as the relatively large stimulus size used by standard perimetry, which exceeds spatial summation. Methods: The relationship between optical coherence tomography (OCT)-derived GCIPL thickness and corresponding inner cell density from retinal flat mounts was determined for four nonhuman primates with varying stages of neuropathy. Normative data for 10-2 threshold using Goldman size I to V stimuli were established for 10 animals, 4 of which were then followed longitudinally with OCT and perimetry. The relationship between GCIPL volume, which incorporated stimulus size after removal of residual thickness, and differential light sensitivity was determined for both experimental glaucoma and healthy eyes. Results: Peak inner retinal cell density was 63,052 ± 9238 cells/mm2 in the healthy eye. Cell density was related to both GCIPL thickness and eccentricity (R2 = 0.74, P < .01). For all 10-2 eccentricities, size III stimuli were greater than the critical area (P < 0.01). Based on the structural and histologic relationship, the critical area corresponds to approximately 156 RGCs. Conclusions: The relationship between cell density and GCIPL thickness is dependent on retinal eccentricity. For 10-2 perimetry, perimetric loss, especially at earlier stages of neuropathy, may best be detected using size II or smaller stimuli.

Transformation-Consistent Self-Ensembling Model for Semisupervised Medical Image Segmentation.

A common shortfall of supervised deep learning for medical imaging is the lack of labeled data, which is often expensive and time consuming to collect. This article presents a new semisupervised method for medical image segmentation, where the network is optimized by a weighted combination of a common supervised loss only for the labeled inputs and a regularization loss for both the labeled and unlabeled data. To utilize the unlabeled data, our method encourages consistent predictions of the network-in-training for the same input under different perturbations. With the semisupervised segmentation tasks, we introduce a transformation-consistent strategy in the self-ensembling model to enhance the regularization effect for pixel-level predictions. To further improve the regularization effects, we extend the transformation in a more generalized form including scaling and optimize the consistency loss with a teacher model, which is an averaging of the student model weights. We extensively validated the proposed semisupervised method on three typical yet challenging medical image segmentation tasks: 1) skin lesion segmentation from dermoscopy images in the International Skin Imaging Collaboration (ISIC) 2017 data set; 2) optic disk (OD) segmentation from fundus images in the Retinal Fundus Glaucoma Challenge (REFUGE) data set; and 3) liver segmentation from volumetric CT scans in the Liver Tumor Segmentation Challenge (LiTS) data set. Compared with state-of-the-art, our method shows superior performance on the challenging 2-D/3-D medical images, demonstrating the effectiveness of our semisupervised method for medical image segmentation.