Automated detection of nine infantile fundus diseases and conditions in retinal images using a deep learning system.

Purpose: We developed an Infant Retinal Intelligent Diagnosis System (IRIDS), an automated system to aid early diagnosis and monitoring of infantile fundus diseases and health conditions to satisfy urgent needs of ophthalmologists. Methods: We developed IRIDS by combining convolutional neural networks and transformer structures, using a dataset of 7697 retinal images (1089 infants) from four hospitals. It identifies nine fundus diseases and conditions, namely, retinopathy of prematurity (ROP) (mild ROP, moderate ROP, and severe ROP), retinoblastoma (RB), retinitis pigmentosa (RP), Coats disease, coloboma of the choroid, congenital retinal fold (CRF), and normal. IRIDS also includes depth attention modules, ResNet-18 (Res-18), and Multi-Axis Vision Transformer (MaxViT). Performance was compared to that of ophthalmologists using 450 retinal images. The IRIDS employed a five-fold cross-validation approach to generate the classification results. Results: Several baseline models achieved the following metrics: accuracy, precision, recall, F1-score (F1), kappa, and area under the receiver operating characteristic curve (AUC) with best values of 94.62% (95% CI, 94.34%-94.90%), 94.07% (95% CI, 93.32%-94.82%), 90.56% (95% CI, 88.64%-92.48%), 92.34% (95% CI, 91.87%-92.81%), 91.15% (95% CI, 90.37%-91.93%), and 99.08% (95% CI, 99.07%-99.09%), respectively. In comparison, IRIDS showed promising results compared to ophthalmologists, demonstrating an average accuracy, precision, recall, F1, kappa, and AUC of 96.45% (95% CI, 96.37%-96.53%), 95.86% (95% CI, 94.56%-97.16%), 94.37% (95% CI, 93.95%-94.79%), 95.03% (95% CI, 94.45%-95.61%), 94.43% (95% CI, 93.96%-94.90%), and 99.51% (95% CI, 99.51%-99.51%), respectively, in multi-label classification on the test dataset, utilizing the Res-18 and MaxViT models. These results suggest that, particularly in terms of AUC, IRIDS achieved performance that warrants further investigation for the detection of retinal abnormalities. Conclusions: IRIDS identifies nine infantile fundus diseases and conditions accurately. It may aid non-ophthalmologist personnel in underserved areas in infantile fundus disease screening. Thus, preventing severe complications. The IRIDS serves as an example of artificial intelligence integration into ophthalmology to achieve better outcomes in predictive, preventive, and personalized medicine (PPPM / 3PM) in the treatment of infantile fundus diseases. Supplementary Information: The online version contains supplementary material available at 10.1007/s13167-024-00350-y.

Effects of newer-generation anti-diabetics on diabetic retinopathy: a critical review.

Diabetic retinopathy (DR) is the leading etiology of blindness in the working population of the USA. Its long-term management relies on effective glycemic control. Seven anti-diabetic classes have been introduced for patients with type 2 diabetes (T2D) in the past two decades, with different glucose-lowering and cardiovascular benefits. Yet, their effects specifically on DR have not been studied in detail. A systematic review of the literature was conducted to investigate this topic, focusing on the available clinical data for T2D. Published studies were evaluated based on their level of statistical evidence, as long as they incorporated at least one endpoint or adverse event pertaining to retinal health. Fifty nine articles met our inclusion criteria and were grouped per anti-diabetic class as follows: alpha-glucosidase inhibitors (1), peroxisome proliferator-activated receptor gamma (PPAR-γ) agonists (8), amylin analogs (1), glucagon-like peptide-1 (GLP-1) receptor agonists (28), dipeptidyl peptidase 4 (DPP-4) inhibitors (9), and sodium glucose co-transporter-2 (SGLT-2) inhibitors (9), plus one retrospective study and two meta-analyses evaluating more than one of the aforementioned anti-diabetic categories. We also reviewed publicly-announced results of trials for the recently-introduced class of twincretins. The available data indicates that most drugs in the newer anti-diabetic classes are neutral to DR progression; however, there are subclasses differences in specific drugs and T2D populations. In particular, there is evidence suggesting there may be worse diabetic macular edema with PPAR-gamma agonists, potential slight DR worsening with semaglutide (GLP-1 receptor agonist), and potential slight increase in the incidence of retinal vein occlusion in elderly and patients with advanced kidney disease receiving SGLT-2 inhibitors. All these warrant further investigation. Longer follow-up and systematic assessment of at least one DR-related endpoint are highly recommended for all future trials in the T2D field, to ultimately address this topic.

An Artificial Intelligence System for Screening and Recommending the Treatment Modalities for Retinopathy of Prematurity.

PURPOSE: The purpose of this study was to develop an artificial intelligence (AI) system for the identification of disease status and recommending treatment modalities for retinopathy of prematurity (ROP). METHODS: This retrospective cohort study included a total of 24,495 RetCam images from 1075 eyes of 651 preterm infants who received RetCam examination at the Shenzhen Eye Hospital in Shenzhen, China, from January 2003 to August 2021. Three tasks included ROP identification, severe ROP identification, and treatment modalities identification (retinal laser photocoagulation or intravitreal injections). The AI system was developed to identify the 3 tasks, especially the treatment modalities of ROP. The performance between the AI system and ophthalmologists was compared using extra 200 RetCam images. RESULTS: The AI system exhibited favorable performance in the 3 tasks, including ROP identification [area under the receiver operating characteristic curve (AUC), 0.9531], severe ROP identification (AUC, 0.9132), and treatment modalities identification with laser photocoagulation or intravitreal injections (AUC, 0.9360). The AI system achieved an accuracy of 0.8627, a sensitivity of 0.7059, and a specificity of 0.9412 for identifying the treatment modalities of ROP. External validation results confirmed the good performance of the AI system with an accuracy of 92.0% in all 3 tasks, which was better than 4 experienced ophthalmologists who scored 56%, 65%, 71%, and 76%, respectively. CONCLUSIONS: The described AI system achieved promising outcomes in the automated identification of ROP severity and treatment modalities. Using such algorithmic approaches as accessory tools in the clinic may improve ROP screening in the future.

The NLRP3 inflammasome - interleukin 1ß axis in uveal melanoma.

Uveal melanoma (UM) is the most common primary intraocular cancer in the adult population. Recent studies suggested that the NLRP3 inflammasome could be a therapeutic target for cutaneous melanoma (CM), but the role of NLRP3 in UM remains unknown. Here, we analyzed the NLRP3-IL-1ß axis in 5 UM and 4 CM cell lines. Expression of NLRP3 mRNA in UM and CM was low, and expression in UM was lower than in CM (P < 0.001). NLRP3 protein levels were below detection limit for all cell lines. UM exhibited lower baseline IL-1ß secretion than CM, especially when compared to the Hs294t cell line (P < 0.05). Bioinformatic analysis of human tumor samples showed that UM has significantly lower expression of NLRP3 and IL-1ß compared with CM. In conclusion, our work shows evidence of extremely low NLRP3 expression and IL-1ß secretion by melanoma cells and highlight differences between CM and UM.

Mesenchymal Stromal/Stem Cell Extracellular Vesicles and Perinatal Injury: One Formula for Many Diseases.

Over the past decades, substantial advances in neonatal medical care have increased the survival of extremely premature infants. However, there continues to be significant morbidity associated with preterm birth with common complications including bronchopulmonary dysplasia (BPD), necrotizing enterocolitis (NEC), neuronal injury such as intraventricular hemorrhage (IVH) or hypoxic ischemic encephalopathy (HIE), as well as retinopathy of prematurity (ROP). Common developmental immune and inflammatory pathways underlie the pathophysiology of such complications providing the opportunity for multisystem therapeutic approaches. To date, no single therapy has proven to be effective enough to prevent or treat the sequelae of prematurity. In the past decade mesenchymal stem/stromal cell (MSC)-based therapeutic approaches have shown promising results in numerous experimental models of neonatal diseases. It is now accepted that the therapeutic potential of MSCs is comprised of their secretome, and several studies have recognized the small extracellular vesicles (sEVs) as the paracrine vector. Herein, we review the current literature on the MSC-EVs as potential therapeutic agents in neonatal diseases and comment on the progress and challenges of their translation to the clinical setting.

Extracellular Vesicles in Corneal Fibrosis/Scarring.

Communication between cells and the microenvironment is a complex, yet crucial, element in the development and progression of varied physiological and pathological processes. Accumulating evidence in different disease models highlights roles of extracellular vesicles (EVs), either in modulating cell signaling paracrine mechanism(s) or harnessing their therapeutic moiety. Of interest, the human cornea functions as a refractive and transparent barrier that protects the intraocular elements from the external environment. Corneal trauma at the ocular surface may lead to diminished corneal clarity and detrimental effects on visual acuity. The aberrant activation of corneal stromal cells, which leads to myofibroblast differentiation and a disorganized extracellular matrix is a central biological process that may result in corneal fibrosis/scarring. In recent years, understanding the pathological and therapeutic EV mechanism(s) of action in the context of corneal biology has been a topic of increasing interest. In this review, we describe the clinical relevance of corneal fibrosis/scarring and how corneal stromal cells contribute to wound repair and their generation of the stromal haze. Furthermore, we will delve into EV characterization, their subtypes, and the pathological and therapeutic roles they play in corneal scarring/fibrosis.

Dissolved oxygen technologies as a novel strategy for non-healing wounds: A critical review.

Non-healing wounds are steadily becoming a global-health issue. Prolonged hypoxia propagates wound chronicity; yet, oxygenating treatments are considered inadequate to date. Dissolved oxygen (DO) in aqueous solutions introduces a novel approach to enhanced wound oxygenation, and is robustly evaluated for clinical applications. A systematic literature search was conducted, whereby experimental and clinical studies of DO technologies were categorized per engineering approach. Technical principles, methodology, endpoints and outcomes were analysed for both oxygenating and healing effects. Forty articles meeting our inclusion criteria were grouped as follows: DO solutions (17), oxygen (O2 ) dressings (9), O2 hydrogels (11) and O2 emulsions (3). All technologies improved wound oxygenation, each to a variable degree. They also achieved at least one statistically significant outcome related to wound healing, mainly in epithelialization, angiogenesis and collagen synthesis. Scarcity in clinical data and methodological variability precluded quantitative comparisons among the biotechnologies studied. DO technologies warrantee further evaluation for wound oxygenation in the clinical setting. Standardised methodologies and targeted research questions are pivotal to facilitate global integration in healthcare.

Incidence of Endophthalmitis after Intravitreal Anti-Vascular Endothelial Growth Factor Injections in an Operating Room in China.

PURPOSE: To evaluate the rate of presumed endophthalmitis (EO) after intravitreal anti-vascular endothelial growth factor (anti-VEGF) injections performed in an operating room (OR) under sterile conditions in mainland China. METHODS: Retrospective single-center study between September 2012 and December 2017 at Beijing Tongren Eye Center, Beijing, China. Intravitreal injection database was reviewed. All anti-VEGF injections were performed using a standardized sterile technique in an OR. Injection protocols included antibiotics for 3 days pre-injection, topical 5% povidone-iodine rinsing before the procedure, and post-injection antibiotics for 3 days. RESULTS: A total of 37,830 intravitreal injections were performed at Beijing Tongren Eye Center. Three cases were managed as presumed EO (0.0079%). Positive cultures were documented in 2 of 3 cases. EO incidence following ranibizumab and conbercept administration was 0.0088% (3 in 33,930) and 0% (0 in 3,900), respectively. No significant difference was detected between the two drugs (P = 0.745). CONCLUSIONS: Very low EO rates were seen in mainland China using a standardized sterile technique in an OR. However, EO could not be completely avoided.