Big data in corneal diseases and cataract: Current applications and future directions

Artificial Intelligence and Ophthalmic Clinical Registries

PURPOSE

The recent advances in artificial intelligence (AI) represent a promising solution to increasing clinical demand and ever limited health resources. Whilst powerful, AI models require vast amounts of representative training data to output meaningful predictions in the clinical environment. Clinical registries represent a promising source of large volume real-world data which could be used to train more accurate and widely applicable AI models. This review aims to provide an overview of the current applications of AI to ophthalmic clinical registry data.

DESIGN AND METHODS

A systematic search of EMBASE, Medline, PubMed, Scopus and Web of Science for primary research articles that applied AI to ophthalmic clinical registry data was conducted in July 2024.

RESULTS

Twenty-three primary research articles applying AI to ophthalmic clinic registries (n = 14) were found. Registries were primarily defined by the condition captured and the most common conditions where AI was applied were glaucoma (n = 3) and neovascular age-related macular degeneration (n = 3). Tabular clinical data was the most common form of input into AI algorithms and outputs were primarily classifiers (n = 8, 40%) and risk quantifier models (n = 7, 35%). The AI algorithms applied were almost exclusively supervised conventional machine learning models (n = 39, 85%) such as decision tree classifiers and logistic regression, with only 7 applications of deep learning or natural language processing algorithms. Significant heterogeneity was found with regards to model validation methodology and measures of performance.

CONCLUSIONS

Limited applications of deep learning algorithms to clinical registry data have been reported. The lack of standardized validation methodology and heterogeneity of performance outcome reporting suggests that the application of AI to clinical registries is still in its infancy constrained by the poor accessibility of registry data and reflecting the need for a standardization of methodology and greater involvement of domain experts in the future development of clinically deployable AI.

Global review of publicly available image datasets for the anterior segment of the eye

This study comprehensively reviewed publicly available image datasets for the anterior segment, with a focus on cataract, refractive, and corneal surgeries. The goal was to assess characteristics of existing datasets and identify areas for improvement. PubMED and Google searches were performed using the search terms "refractive surgery," "anterior segment," "cornea," "corneal," "cataract" AND "database," with the related word of "imaging." Results of each of these searches were collated, identifying 26 publicly available anterior segment image datasets. Imaging modalities included optical coherence tomography, photography, and confocal microscopy. Most datasets were small, 80% originated in the U.S., China, or Europe. Over 50% of images were from normal eyes. Disease states represented included keratoconus, corneal ulcers, and Fuchs dystrophy. Most of the datasets were incompletely described. To promote accessibility going forward to 2030, the ESCRS Digital Health Special Interest Group will annually update a list of available image datasets for anterior segment at www.escrs.org .

Analyzing the changing trend of corneal biomechanical properties under different influencing factors in T2DM patients

To analyze the changing trend of CH and CRF values under different influencing factors in T2DM patients. A total of 650 patients with T2DM were included. We discovered that the course of T2DM, smoking history, BMI, and FBG, DR, HbA1c, TC, TG, and LDL-C levels were common risk factors for T2DM, while HDL-C levels were a protective factor. Analyzing the CH and CRF values according to the course of diabetes, we discovered that as T2DM continued to persist, the values of CH and CRF gradually decreased. Moreover, with the increase in FBG levels and the accumulation of HbA1c, the values of CH and CRF gradually decreased. In addition, in patients with HbA1c (%) > 12, the values of CH and CRF decreased the most, falling by 1.85 ± 0.33 mmHg and 1.28 ± 0.69 mmHg, respectively. Compared with the non-DR group, the CH and CRF values gradually decreased in the mild-NPDR, moderate-NPDR, severe-NPDR and PDR groups, with the lowest CH and CRF values in the PDR group. In patients with T2DM, early measurement of corneal biomechanical properties to evaluate the change trend of CH and CRF values in different situations will help to identify and prevent diabetic keratopathy in a timely manner.

Clinical Applications of Anterior Segment Optical Coherence Tomography: An Updated Review

Since its introduction, optical coherence tomography (OCT) has revolutionized the field of ophthalmology and has now become an indispensable, noninvasive tool in daily practice. Most ophthalmologists are familiar with its use in the assessment and monitoring of retinal and optic nerve diseases. However, it also has important applications in the assessment of anterior segment structures, including the cornea, conjunctiva, sclera, anterior chamber, and iris, and has the potential to transform the clinical examination of these structures. In this review, we aim to provide a comprehensive overview of the potential clinical utility of anterior segment OCT (AS-OCT) for a wide range of anterior segment pathologies, such as conjunctival neoplasia, pterygium, scleritis, keratoconus, corneal dystrophies, and infectious/noninfectious keratitis. In addition, the clinical applications of AS-OCT (including epithelial mapping) in preoperative planning and postoperative monitoring for corneal and refractive surgeries are discussed.

Cataract surgery after corneal refractive surgery: preoperative considerations and management

PURPOSE OF REVIEW

Corneal refractive surgery (CRS) is one of the most popular eye procedures, with more than 40 million cases performed globally. As CRS-treated patients age and develop cataract, the number of cases that require additional preoperative considerations and management will increase around the world. Thus, we provide an up-to-date, concise overview of the considerations and outcomes of cataract surgery in eyes with previous CRS, including surface ablation, laser in-situ keratomileusis (LASIK), and small-incision lenticule extraction (SMILE).

RECENT FINDINGS

Challenges associated with accurate biometry in eyes with CRS have been mitigated recently through total keratometry, ray tracing, intraoperative aberrometry, and machine learning assisted intraocular lens (IOL) power calculation formulas to improve prediction. Emerging studies have highlighted the superior performance of ray tracing and/or total keratometry-based formulas for IOL power calculation in eyes with previous SMILE. Dry eye remains a common side effect after cataract surgery, especially in eyes with CRS, though the risk appears to be lower after SMILE than LASIK (in the short-term). Recent presbyopia-correcting IOL designs such as extended depth of focus (EDOF) IOLs may be suitable in carefully selected eyes with previous CRS.

SUMMARY

Ophthalmologists will increasingly face challenges associated with the surgical management of cataract in patients with prior CRS. Careful preoperative assessment of the ocular surface, appropriate use of IOL power calculation formulas, and strategies for presbyopia correction are key to achieve good clinical and refractive outcomes and patient satisfaction. Recent advances in CRS techniques, such as SMILE, may pose new challenges for such eyes in the future.

The Royal College of Ophthalmologists' National Ophthalmology Database study of cataract surgery: report 16, influence of remuneration model on choice of intraocular lens in the UK

BACKGROUND/OBJECTIVES

Cataract surgery with intraocular lens (IOL) implantation is one of the most commonly performed surgeries worldwide. Within the UK, publicly funded cataract surgery is remunerated by two models: (1) "block contract" (BC), which commissions organisations to deliver whole service pathways without considering specific activity items; or (2) "payment by results" (PbR), which pays a tariff price for each procedure. This study aimed to examine the association between remuneration model and the cost and types of IOL used.

SUBJECTS/METHODS

Cataract operations recorded on the Royal College of Ophthalmologists' National Ophthalmology Database were included, with additional data collected for remuneration model from NHS England and cost of IOL from the NHS Spend Comparison Service.

RESULTS

We included 907,052 cataract operations from 87 centres. The majority of operations were performed in PbR centres (456 198, 50.3%), followed by BC centres (240 641, 26.5%) and mixed models centres (210 213, 23.2%). The mean price of hydrophobic (n = 7) and hydrophilic IOLs (n = 5) were £45.72 and £42.86, respectively. Hydrophobic IOLs were predominantly used (650 633, 71.7%) and were significantly more commonly used in centres remunerated by BC (96.5% vs. 3.5%) than those by PbR (65.7% vs. 34.3%) when compared to hydrophilic IOLs (p < 0.001).

CONCLUSIONS

This study demonstrated that the IOL choice may be perversely incentivised by the IOL cost and remuneration model. Although hydrophobic IOLs are more expensive at the point of surgery, their potential longer-term cost-effectiveness due to reduced requirement for YAG capsulotomy should be considered.

Animal Models in Eye Research: Focus on Corneal Pathologies

The eye is a complex sensory organ that enables visual perception of the world. The dysfunction of any of these tissues can impair vision. Conduction studies on laboratory animals are essential to ensure the safety of therapeutic products directly applied or injected into the eye to treat ocular diseases before eventually proceeding to clinical trials. Among these tissues, the cornea has unique homeostatic and regenerative mechanisms for maintaining transparency and refraction of external light, which are essential for vision. However, being the outermost tissue of the eye and directly exposed to the external environment, the cornea is particularly susceptible to injury and diseases. This review highlights the evidence for selecting appropriate animals to better understand and treat corneal diseases, which rank as the fifth leading cause of blindness worldwide. The development of reliable and human-relevant animal models is, therefore, a valuable research tool for understanding and translating fundamental mechanistic findings, as well as for assessing therapeutic potential in humans. First, this review emphasizes the unique characteristics of animal models used in ocular research. Subsequently, it discusses current animal models associated with human corneal pathologies, their utility in understanding ocular disease mechanisms, and their role as translational models for patients.

A 7-year review of clinical characteristics, predisposing factors and outcomes of post-keratoplasty infectious keratitis: the Nottingham infectious keratitis study

Background/objectives

Post-keratoplasty infectious keratitis (PKIK) is a unique sight-threatening clinical entity which often poses significant therapeutic challenges. This study aimed to examine the clinical presentation, risk factors, management, and clinical outcomes of PKIK.

Methods

This was a retrospective study of all patients who presented to the Queen's Medical Centre, Nottingham, with PKIK between September 2015 and August 2022 (a 7-year period). Relevant data on types of keratoplasty, clinical presentations, causative microorganisms, management, and outcome were analyzed.

Results

Forty-nine PKIK cases, including four cases of interface infectious keratitis, were identified during the study period. The most common graft indications for PKP, DALK and EK were failed grafts (9, 37.5%), keratoconus (6, 54.5%) and Fuchs endothelial corneal dystrophy (FECD; 8, 57.1%), respectively. Staphylococcus spp. were the most commonly identified organisms (15, 50.0%). Bullous keratopathy (18, 36.7%), ocular surface disease (18, 36.7%), and broken/loose sutures (15, 30.6%) were the most common risk factors. Concurrent use of topical steroids was identified in 25 (51.0%) cases. Of 31 functioning grafts at presentation, 12 (38.7%) grafts failed at final follow-up with 15 (48.4%) patients retaining a CDVA of ≥1.0 logMAR. The overall estimated 5-year survival rate post-PKIK was 55.9% (95% CI, 35.9%-75.9%), with DALK having the highest survival rate [63.6% (95% CI, 28.9%-98.3%)], followed by EK [57.1% (95% CI, 20.4%-93.8%)] and PKP [52.7% (95% CI, 25.1%-80.3%)], though no statistical difference was observed (p=0.48).

Conclusions

PKIK represents an important cause of IK and graft failure. Bullous keratopathy, OSD and suture-related complications are the commonest risk factors, highlighting the potential benefit of prophylactic topical antibiotics (for unhealthy ocular surface) and early suture removal (where possible) in reducing the risk of PKIK. Graft survival may be higher in lamellar keratoplasty following PKIK but larger studies are required to elucidate this observation.

Diagnosis of Acanthamoeba Keratitis: Past, Present and Future

Acanthamoeba keratitis (AK) is a painful and sight-threatening parasitic corneal infection. In recent years, the incidence of AK has increased. Timely and accurate diagnosis is crucial during the management of AK, as delayed diagnosis often results in poor clinical outcomes. Currently, AK diagnosis is primarily achieved through a combination of clinical suspicion, microbiological investigations and corneal imaging. Historically, corneal scraping for microbiological culture has been considered to be the gold standard. Despite its technical ease, accessibility and cost-effectiveness, the long diagnostic turnaround time and variably low sensitivity of microbiological culture limit its use as a sole diagnostic test for AK in clinical practice. In this review, we aim to provide a comprehensive overview of the diagnostic modalities that are currently used to diagnose AK, including microscopy with staining, culture, corneal biopsy, in vivo confocal microscopy, polymerase chain reaction and anterior segment optical coherence tomography. We also highlight emerging techniques, such as next-generation sequencing and artificial intelligence-assisted models, which have the potential to transform the diagnostic landscape of AK.

Management of keratoconus: an updated review

Keratoconus is the most common corneal ectatic disorder. It is characterized by progressive corneal thinning with resultant irregular astigmatism and myopia. Its prevalence has been estimated at 1:375 to 1:2,000 people globally, with a considerably higher rate in the younger populations. Over the past two decades, there was a paradigm shift in the management of keratoconus. The treatment has expanded significantly from conservative management (e.g., spectacles and contact lenses wear) and penetrating keratoplasty to many other therapeutic and refractive modalities, including corneal cross-linking (with various protocols/techniques), combined CXL-keratorefractive surgeries, intracorneal ring segments, anterior lamellar keratoplasty, and more recently, Bowman's layer transplantation, stromal keratophakia, and stromal regeneration. Several recent large genome-wide association studies (GWAS) have identified important genetic mutations relevant to keratoconus, facilitating the development of potential gene therapy targeting keratoconus and halting the disease progression. In addition, attempts have been made to leverage the power of artificial intelligence-assisted algorithms in enabling earlier detection and progression prediction in keratoconus. In this review, we provide a comprehensive overview of the current and emerging treatment of keratoconus and propose a treatment algorithm for systematically guiding the management of this common clinical entity.

Diagnostic performance of deep learning in infectious keratitis: a systematic review and meta-analysis protocol

INTRODUCTION

Infectious keratitis (IK) represents the fifth-leading cause of blindness worldwide. A delay in diagnosis is often a major factor in progression to irreversible visual impairment and/or blindness from IK. The diagnostic challenge is further compounded by low microbiological culture yield, long turnaround time, poorly differentiated clinical features and polymicrobial infections. In recent years, deep learning (DL), a subfield of artificial intelligence, has rapidly emerged as a promising tool in assisting automated medical diagnosis, clinical triage and decision-making, and improving workflow efficiency in healthcare services. Recent studies have demonstrated the potential of using DL in assisting the diagnosis of IK, though the accuracy remains to be elucidated. This systematic review and meta-analysis aims to critically examine and compare the performance of various DL models with clinical experts and/or microbiological results (the current 'gold standard') in diagnosing IK, with an aim to inform practice on the clinical applicability and deployment of DL-assisted diagnostic models.

METHODS AND ANALYSIS

This review will consider studies that included application of any DL models to diagnose patients with suspected IK, encompassing bacterial, fungal, protozoal and/or viral origins. We will search various electronic databases, including EMBASE and MEDLINE, and trial registries. There will be no restriction to the language and publication date. Two independent reviewers will assess the titles, abstracts and full-text articles. Extracted data will include details of each primary studies, including title, year of publication, authors, types of DL models used, populations, sample size, decision threshold and diagnostic performance. We will perform meta-analyses for the included primary studies when there are sufficient similarities in outcome reporting.

ETHICS AND DISSEMINATION

No ethical approval is required for this systematic review. We plan to disseminate our findings via presentation/publication in a peer-reviewed journal.

PROSPERO REGISTRATION NUMBER

CRD42022348596.