Artificial Intelligence for Cataract Detection and Management

Assessing the Impact of AI-Assisted Portable Slit Lamps on Rural Primary Ophthalmic Medical Service

PURPOSE

To investigate the effect of an AI-assisted portable slit lamp (iSpector) and basic ophthalmology training on cataract detection, referral, and surgery rate in rural areas.

METHODS

This randomized control trial randomly assigned 63 village doctors to either the AI-assisted group (providing iSpector and training) or the control group (providing training). Doctors were followed for 1 year before intervention as a baseline and 1 year after to make the comparison. Multivariable Poisson regression was applied to compare the difference in cataract detection, referral, and surgery rate between the two groups, adjusted for primary doctors' baseline characteristics. We further conducted subgroup analysis to estimate the change after the intervention.

RESULTS

Compared to the control group, the detection, referral, and surgery rate of cataracts among the AI-assisted group was comparable, 1.7 times higher, and 4.9 times higher, respectively. Providing iSpector and training increased the detection, referral, and surgery rate of cataracts. However, only based on training to elevate the detection rate of cataracts did not change the referral and surgery rate.

CONCLUSIONS

iSpector helps village doctors detect and refer cataract patients appropriately, thus increasing the probability that patients receive cataract surgery.

Slit Lamp Report Generation and Question Answering: Development and Validation of a Multimodal Transformer Model with Large Language Model Integration

BACKGROUND

Large language models have shown remarkable efficacy in various medical research and clinical applications. However, their skills in medical image recognition and subsequent report generation or question answering (QA) remain limited.

OBJECTIVE

We aim to finetune a multimodal, transformer-based model for generating medical reports from slit lamp images and develop a QA system using Llama2. We term this entire process slit lamp-GPT.

METHODS

Our research used a dataset of 25,051 slit lamp images from 3409 participants, paired with their corresponding physician-created medical reports. We used these data, split into training, validation, and test sets, to finetune the Bootstrapping Language-Image Pre-training framework toward report generation. The generated text reports and human-posed questions were then input into Llama2 for subsequent QA. We evaluated performance using qualitative metrics (including BLEU [bilingual evaluation understudy], CIDEr [consensus-based image description evaluation], ROUGE-L [Recall-Oriented Understudy for Gisting Evaluation-Longest Common Subsequence], SPICE [Semantic Propositional Image Caption Evaluation], accuracy, sensitivity, specificity, precision, and F1-score) and the subjective assessments of two experienced ophthalmologists on a 1-3 scale (1 referring to high quality).

RESULTS

We identified 50 conditions related to diseases or postoperative complications through keyword matching in initial reports. The refined slit lamp-GPT model demonstrated BLEU scores (1-4) of 0.67, 0.66, 0.65, and 0.65, respectively, with a CIDEr score of 3.24, a ROUGE (Recall-Oriented Understudy for Gisting Evaluation) score of 0.61, and a Semantic Propositional Image Caption Evaluation score of 0.37. The most frequently identified conditions were cataracts (22.95%), age-related cataracts (22.03%), and conjunctival concretion (13.13%). Disease classification metrics demonstrated an overall accuracy of 0.82 and an F1-score of 0.64, with high accuracies (≥0.9) observed for intraocular lens, conjunctivitis, and chronic conjunctivitis, and high F1-scores (≥0.9) observed for cataract and age-related cataract. For both report generation and QA components, the two evaluating ophthalmologists reached substantial agreement, with κ scores between 0.71 and 0.84. In assessing 100 generated reports, they awarded scores of 1.36 for both completeness and correctness; 64% (64/100) were considered "entirely good," and 93% (93/100) were "acceptable." In the evaluation of 300 generated answers to questions, the scores were 1.33 for completeness, 1.14 for correctness, and 1.15 for possible harm, with 66.3% (199/300) rated as "entirely good" and 91.3% (274/300) as "acceptable."

CONCLUSIONS

This study introduces the slit lamp-GPT model for report generation and subsequent QA, highlighting the potential of large language models to assist ophthalmologists and patients.

Acceptance of AI in Health Care for Short- and Long-Term Treatments: Pilot Development Study of an Integrated Theoretical Model

BACKGROUND

As digital technologies and especially artificial intelligence (AI) become increasingly important in health care, it is essential to determine whether and why potential users intend to use related health information systems (HIS). Several theories exist, but they focus mainly on aspects of health care or information systems, in addition to general psychological theories, and hence provide a small number of variables to explain future behavior. Thus, research that provides a larger number of variables by combining several theories from health care, information systems, and psychology is necessary.

OBJECTIVE

This study aims to investigate the intention to use new HIS for decisions concerning short- and long-term medical treatments using an integrated approach with several variables to explain future behavior.

METHODS

We developed an integrated theoretical model based on theories from health care, information systems, and psychology that allowed us to analyze the duality approach of adaptive and nonadaptive appraisals and their influence on the intention to use HIS. We applied the integrated theoretical model to the short-term treatment using AI-based HIS for surgery and the long-term treatment of diabetes tracking using survey data with structured equation modeling. To differentiate between certain levels of AI involvement, we used several scenarios that include treatments by physicians only, physicians with AI support, and AI only to understand how individuals perceive the influence of AI.

RESULTS

Our results showed that for short- and long-term treatments, the variables perceived threats, fear (disease), perceived efficacy, attitude (HIS), and perceived norms are important to consider when determining the intention to use AI-based HIS. Furthermore, the results revealed that perceived efficacy and attitude (HIS) are the most important variables to determine intention to use for all treatments and scenarios. In contrast, abilities (HIS) were important for short-term treatments only. For our 9 scenarios, adaptive and nonadaptive appraisals were both important to determine intention to use, depending on whether the treatment is known. Furthermore, we determined R² values that varied between 57.9% and 81.7% for our scenarios, which showed that the explanation power of our model is medium to good.

CONCLUSIONS

We contribute to HIS literature by highlighting the importance of integrating disease- and technology-related factors and by providing an integrated theoretical model. As such, we show how adaptive and nonadaptive appraisals should be arranged to report on medical decisions in the future, especially in the short and long terms. Physicians and HIS developers can use our insights to identify promising rationale for HIS adoption concerning short- and long-term treatments and adapt and develop HIS accordingly. Specifically, HIS developers should ensure that future HIS act in terms of HIS functions, as our study shows that efficient HIS lead to a positive attitude toward the HIS and ultimately to a higher intention to use.

S, Kochar P, Selvaraj S, Levine H, Amescua G, Sepulveda-Beltran PA, Niziol LM, Woodward MA. Automatic Classification of Slit-Lamp Photographs by Imaging Illumination

PURPOSE

The aim of this study was to facilitate deep learning systems in image annotations for diagnosing keratitis type by developing an automated algorithm to classify slit-lamp photographs (SLPs) based on illumination technique.

METHODS

SLPs were collected from patients with corneal ulcer at Kellogg Eye Center, Bascom Palmer Eye Institute, and Aravind Eye Care Systems. Illumination techniques were slit beam, diffuse white light, diffuse blue light with fluorescein, and sclerotic scatter (ScS). Images were manually labeled for illumination and randomly split into training, validation, and testing data sets (70%:15%:15%). Classification algorithms including MobileNetV2, ResNet50, LeNet, AlexNet, multilayer perceptron, and k-nearest neighborhood were trained to distinguish 4 type of illumination techniques. The algorithm performances on the test data set were evaluated with 95% confidence intervals (CIs) for accuracy, F1 score, and area under the receiver operator characteristics curve (AUC-ROC), overall and by class (one-vs-rest).

RESULTS

A total of 12,132 images from 409 patients were analyzed, including 41.8% (n = 5069) slit-beam photographs, 21.2% (2571) diffuse white light, 19.5% (2364) diffuse blue light, and 17.5% (2128) ScS. MobileNetV2 achieved the highest overall F1 score of 97.95% (CI, 97.94%-97.97%), AUC-ROC of 99.83% (99.72%-99.9%), and accuracy of 98.98% (98.97%-98.98%). The F1 scores for slit beam, diffuse white light, diffuse blue light, and ScS were 97.82% (97.80%-97.84%), 96.62% (96.58%-96.66%), 99.88% (99.87%-99.89%), and 97.59% (97.55%-97.62%), respectively. Slit beam and ScS were the 2 most frequently misclassified illumination.

CONCLUSIONS

MobileNetV2 accurately labeled illumination of SLPs using a large data set of corneal images. Effective, automatic classification of SLPs is key to integrating deep learning systems for clinical decision support into practice workflows.

Enhancing Vault Prediction and ICL Sizing Through Advanced Machine Learning Models

PURPOSE

To use artificial intelligence (AI) technology to accurately predict vault and Implantable Collamer Lens (ICL) size.

METHODS

The methodology focused on enhancing predictive capabilities through the fusion of machine-learning algorithms. Specifically, AdaBoost, Random Forest, Decision Tree, Support Vector Regression, LightGBM, and XGBoost were integrated into a majority-vote model. The performance of each model was evaluated using appropriate metrics such as accuracy, precision, F1-score, and area under the curve (AUC).

RESULTS

The majority-vote model exhibited the highest performance among the classification models, with an accuracy of 81.9% area under the curve (AUC) of 0.807. Notably, LightGBM (accuracy = 0.788, AUC = 0.803) and XGBoost (ACC = 0.790, AUC = 0.801) demonstrated competitive results. For the ICL size prediction, the Random Forest model achieved an impressive accuracy of 85.3% (AUC = 0.973), whereas XG-Boost (accuracy = 0.834, AUC = 0.961) and LightGBM (accuracy = 0.816, AUC = 0.961) maintained their compatibility.

CONCLUSIONS

This study highlights the potential of diverse machine learning algorithms to enhance postoperative vault and ICL size prediction, ultimately contributing to the safety of ICL implantation procedures. Furthermore, the introduction of the novel majority-vote model demonstrates its capability to combine the advantages of multiple models, yielding superior accuracy. Importantly, this study will empower ophthalmologists to use a precise tool for vault prediction, facilitating informed ICL size selection in clinical practice. [J Refract Surg. 2024;40(3):e126-e132.].

Smartphone Eye Examination: Artificial Intelligence and Telemedicine

Background: The current medical scenario is closely linked to recent progress in telecommunications, photodocumentation, and artificial intelligence (AI). Smartphone eye examination may represent a promising tool in the technological spectrum, with special interest for primary health care services. Obtaining fundus imaging with this technique has improved and democratized the teaching of fundoscopy, but in particular, it contributes greatly to screening diseases with high rates of blindness. Eye examination using smartphones essentially represents a cheap and safe method, thus contributing to public policies on population screening. This review aims to provide an update on the use of this resource and its future prospects, especially as a screening and ophthalmic diagnostic tool. Methods: In this review, we surveyed major published advances in retinal and anterior segment analysis using AI. We performed an electronic search on the Medical Literature Analysis and Retrieval System Online (MEDLINE), EMBASE, and Cochrane Library for published literature without a deadline. We included studies that compared the diagnostic accuracy of smartphone ophthalmoscopy for detecting prevalent diseases with an accurate or commonly employed reference standard. Results: There are few databases with complete metadata, providing demographic data, and few databases with sufficient images involving current or new therapies. It should be taken into consideration that these are databases containing images captured using different systems and formats, with information often being excluded without essential detailing of the reasons for exclusion, which further distances them from real-life conditions. The safety, portability, low cost, and reproducibility of smartphone eye images are discussed in several studies, with encouraging results. Conclusions: The high level of agreement between conventional and a smartphone method shows a powerful arsenal for screening and early diagnosis of the main causes of blindness, such as cataract, glaucoma, diabetic retinopathy, and age-related macular degeneration. In addition to streamlining the medical workflow and bringing benefits for public health policies, smartphone eye examination can make safe and quality assessment available to the population.

The Current State of Artificial Intelligence in Neuro-Ophthalmology. A Review

This article presents a summary of recent advances in the development and use of complex systems using artificial intelligence (AI) in neuro-ophthalmology. The aim of the following article is to present the principles of AI and algorithms that are currently being used or are still in the stage of evaluation or validation within the neuro-ophthalmology environment. For the purpose of this text, a literature search was conducted using specific keywords in available scientific databases, cumulatively up to April 2023. The AI systems developed across neuro-ophthalmology mostly achieve high sensitivity, specificity and accuracy. Individual AI systems and algorithms are subsequently selected, simply described and compared in the article. The results of the individual studies differ significantly, depending on the chosen methodology, the set goals, the size of the test, evaluated set, and the evaluated parameters. It has been demonstrated that the evaluation of various diseases will be greatly speeded up with the help of AI and make the diagnosis more efficient in the future, thus showing a high potential to be a useful tool in clinical practice even with a significant increase in the number of patients.

AI-based diagnosis of nuclear cataract from slit-lamp videos

In ophthalmology, the availability of many fundus photographs and optical coherence tomography images has spurred consideration of using artificial intelligence (AI) for diagnosing retinal and optic nerve disorders. However, AI application for diagnosing anterior segment eye conditions remains unfeasible due to limited standardized images and analysis models. We addressed this limitation by augmenting the quantity of standardized optical images using a video-recordable slit-lamp device. We then investigated whether our proposed machine learning (ML) AI algorithm could accurately diagnose cataracts from videos recorded with this device. We collected 206,574 cataract frames from 1812 cataract eye videos. Ophthalmologists graded the nuclear cataracts (NUCs) using the cataract grading scale of the World Health Organization. These gradings were used to train and validate an ML algorithm. A validation dataset was used to compare the NUC diagnosis and grading of AI and ophthalmologists. The results of individual cataract gradings were: NUC 0: area under the curve (AUC) = 0.967; NUC 1: AUC = 0.928; NUC 2: AUC = 0.923; and NUC 3: AUC = 0.949. Our ML-based cataract diagnostic model achieved performance comparable to a conventional device, presenting a promising and accurate auto diagnostic AI tool.

Comprehensive Review on the Use of Artificial Intelligence in Ophthalmology and Future Research Directions

BACKGROUND

Having several applications in medicine, and in ophthalmology in particular, artificial intelligence (AI) tools have been used to detect visual function deficits, thus playing a key role in diagnosing eye diseases and in predicting the evolution of these common and disabling diseases. AI tools, i.e., artificial neural networks (ANNs), are progressively involved in detecting and customized control of ophthalmic diseases. The studies that refer to the efficiency of AI in medicine and especially in ophthalmology were analyzed in this review.

MATERIALS AND METHODS

We conducted a comprehensive review in order to collect all accounts published between 2015 and 2022 that refer to these applications of AI in medicine and especially in ophthalmology. Neural networks have a major role in establishing the demand to initiate preliminary anti-glaucoma therapy to stop the advance of the disease.

RESULTS

Different surveys in the literature review show the remarkable benefit of these AI tools in ophthalmology in evaluating the visual field, optic nerve, and retinal nerve fiber layer, thus ensuring a higher precision in detecting advances in glaucoma and retinal shifts in diabetes. We thus identified 1762 applications of artificial intelligence in ophthalmology: review articles and research articles (301 pub med, 144 scopus, 445 web of science, 872 science direct). Of these, we analyzed 70 articles and review papers (diabetic retinopathy (N = 24), glaucoma (N = 24), DMLV (N = 15), other pathologies (N = 7)) after applying the inclusion and exclusion criteria.

CONCLUSION

In medicine, AI tools are used in surgery, radiology, gynecology, oncology, etc., in making a diagnosis, predicting the evolution of a disease, and assessing the prognosis in patients with oncological pathologies. In ophthalmology, AI potentially increases the patient's access to screening/clinical diagnosis and decreases healthcare costs, mainly when there is a high risk of disease or communities face financial shortages. AI/DL (deep learning) algorithms using both OCT and FO images will change image analysis techniques and methodologies. Optimizing these (combined) technologies will accelerate progress in this area.

Health Economic Implications of Artificial Intelligence Implementation for Ophthalmology in Australia: A Systematic Review

PURPOSE

The health care industry is an inherently resource-intense sector. Emerging technologies such as artificial intelligence (AI) are at the forefront of advancements in health care. The health economic implications of this technology have not been clearly established and represent a substantial barrier to adoption both in Australia and globally. This review aims to determine the health economic impact of implementing AI to ophthalmology in Australia.

METHODS

A systematic search of the databases PubMed/MEDLINE, EMBASE, and CENTRAL was conducted to March 2022, before data collection and risk of bias analysis in accordance with preferred reporting items for systematic ceviews and meta-analyses 2020 guidelines (PROSPERO number CRD42022325511). Included were full-text primary research articles analyzing a population of patients who have or are being evaluated for an ophthalmological diagnosis, using a health economic assessment system to assess the cost-effectiveness of AI.

RESULTS

Seven articles were identified for inclusion. Economic viability was defined as direct cost to the patient that is equal to or less than costs incurred with human clinician assessment. Despite the lack of Australia-specific data, foreign analyses overwhelmingly showed that AI is just as economically viable, if not more so, than traditional human screening programs while maintaining comparable clinical effectiveness. This evidence was largely in the setting of diabetic retinopathy screening.

CONCLUSIONS

Primary Australian research is needed to accurately analyze the health economic implications of implementing AI on a large scale. Further research is also required to analyze the economic feasibility of adoption of AI technology in other areas of ophthalmology, such as glaucoma and cataract screening.

Machine Learning for Cataract Classification/Grading on Ophthalmic Imaging Modalities: A Survey

Cataracts are the leading cause of visual impairment and blindness globally. Over the years, researchers have achieved significant progress in developing state-of-the-art machine learning techniques for automatic cataract classification and grading, aiming to prevent cataracts early and improve clinicians’ diagnosis efficiency. This survey provides a comprehensive survey of recent advances in machine learning techniques for cataract classification/grading based on ophthalmic images. We summarize existing literature from two research directions: conventional machine learning methods and deep learning methods. This survey also provides insights into existing works of both merits and limitations. In addition, we discuss several challenges of automatic cataract classification/grading based on machine learning techniques and present possible solutions to these challenges for future research.

Characterization of Dysfunctional Lens Index and Opacity Grade in a Healthy Population

This study enrolled 61 volunteers (102 eyes) classified into subjects < 50 years (group 1) and subjects ≥ 50 years (group 2). Dysfunctional Lens Index (DLI); opacity grade; pupil diameter; and corneal, internal, and ocular higher order aberrations (HOAs) were measured with the i-Trace system (Tracey Technologies). Mean DLI was 8.89 ± 2.00 and 6.71 ± 2.97 in groups 1 and 2, respectively, being significantly higher in group 1 in all and right eyes (both p < 0.001). DLI correlated significantly with age (Rho = −0.41, p < 0.001) and pupil diameter (Rho = 0.20, p = 0.043) for all eyes, and numerous internal and ocular root-mean square HOAs for right, left, and all eyes (Rho ≤ −0.25, p ≤ 0.001). Mean opacity grade was 1.21 ± 0.63 and 1.48 ± 1.15 in groups 1 and 2, respectively, with no significant differences between groups (p ≥ 0.29). Opacity grade significantly correlated with pupil diameter for right and all eyes (Rho ≤ 0.33, p ≤ 0.013), and with some ocular root-mean square HOAs for right and all eyes (Rho ≥ 0.23, p ≤ 0.020). DLI correlates with age and might be used complementary to other diagnostic measurements for assessing the dysfunctional lens syndrome. Both DLI and opacity grade maintain a relationship with pupil diameter and internal and ocular HOAs, supporting that the algorithms used by the device may be based, in part, on these parameters.

Survey-based Evaluation of the Use of Picture Archiving and Communication Systems in an Eye Hospital-Ophthalmologists' Perspective

PURPOSE

Picture archiving and communication system (PACS) is a medical imaging system for sharing, storage, retrieval, and access of medical images stored. Our study aimed to identify ophthalmologists' views on PACS, with the comparison between 3 platforms, namely electronic patient record (ePR), HEYEX (Heidelberg Engineering, Switzerland), and FORUM (Zeiss, US), following their implementation in an eye hospital for common ophthalmic investigations [visual field, optical coherence tomography (OCT) of retinal nerve fiber layer and macula, and fluorescein/indocyanine green angiography (FA/ICG)].

METHODS

An online survey was distributed among ophthalmologists in a single center. Primary outcome included comparison of PACS with paper-based system. Secondary outcomes included pattern of use and comparison of different PACS platforms.

RESULTS

Survey response rate was 28/37 (75.7%). Images were most commonly accessed through ePR (median: 80% of time, interquartile range: 50 to 90%).All systems scored highly in information display items (median scores ≥7.5 out of 10) and in reducing patient identification error in investigation filing and retrieval during consultation compared to paper (score ≥7.0). However, ePR was inferior to paper in "facilitating comparison with previous results" in all investigation types (scores 3.0 to 4.5). ePR scored significantly higher in all system quality items than HEYEX ( P  < 0.001) and FORUM ( P  < 0.022), except login response time ( P  = 0.081). HEYEX scored significantly higher among vitreoretinaluveitis members (VRU) for information quality items for OCT macula and FA/ICG [VRU: 10.0 (8.0 to 10.0), non-VRU: 8.0 (6.75 to 9.25), P  = 0.042].

CONCLUSIONS

Overall feedback for PACS among ophthalmologists was positive, with limitations of inefficiency in use of information, for example, comparison with previous results. Subspecialty played an important role in evaluating PACS.

Digital Education in Ophthalmology

ABSTRACT

Accessibility to the Internet and computer systems has prompted the gravitation towards digital learning in medicine, including ophthalmology. Using the PubMed database and Google search engine, current initiatives in ophthalmology that serve as alternatives to traditional in-person learning with the purpose of enhancing clinical and surgical training were reviewed. This includes the development of teleeducation modules, construction of libraries of clinical and surgical videos, conduction of didactics via video communication, and the implementation of simulators and intelligent tutoring systems into clinical and surgical training programs. In this age of digital communication, teleophthalmology programs, virtual ophthalmological society meetings, and online examinations have become necessary for conducting clinical work and educational training in ophthalmology, especially in light of recent global events that have prevented large gatherings as well as the rural location of various populations. Looking forward, web-based modules and resources, artificial intelligence-based systems, and telemedicine programs will augment current curricula for ophthalmology trainees.

Diagnostic accuracy of code-free deep learning for detection and evaluation of posterior capsule opacification

OBJECTIVE

To train and validate a code-free deep learning system (CFDLS) on classifying high-resolution digital retroillumination images of posterior capsule opacification (PCO) and to discriminate between clinically significant and non-significant PCOs.

METHODS AND ANALYSIS

For this retrospective registry study, three expert observers graded two independent datasets of 279 images three separate times with no PCO to severe PCO, providing binary labels for clinical significance. The CFDLS was trained and internally validated using 179 images of a training dataset and externally validated with 100 images. Model development was through Google Cloud AutoML Vision. Intraobserver and interobserver variabilities were assessed using Fleiss kappa (κ) coefficients and model performance through sensitivity, specificity and area under the curve (AUC).

RESULTS

Intraobserver variability κ values for observers 1, 2 and 3 were 0.90 (95% CI 0.86 to 0.95), 0.94 (95% CI 0.90 to 0.97) and 0.88 (95% CI 0.82 to 0.93). Interobserver agreement was high, ranging from 0.85 (95% CI 0.79 to 0.90) between observers 1 and 2 to 0.90 (95% CI 0.85 to 0.94) for observers 1 and 3. On internal validation, the AUC of the CFDLS was 0.99 (95% CI 0.92 to 1.0); sensitivity was 0.89 at a specificity of 1. On external validation, the AUC was 0.97 (95% CI 0.93 to 0.99); sensitivity was 0.84 and specificity was 0.92.

CONCLUSION

This CFDLS provides highly accurate discrimination between clinically significant and non-significant PCO equivalent to human expert graders. The clinical value as a potential decision support tool in different models of care warrants further research.

Detecting visually significant cataract using retinal photograph-based deep learning

Age-related cataracts are the leading cause of visual impairment among older adults. Many significant cases remain undiagnosed or neglected in communities, due to limited availability or accessibility to cataract screening. In the present study, we report the development and validation of a retinal photograph-based, deep-learning algorithm for automated detection of visually significant cataracts, using more than 25,000 images from population-based studies. In the internal test set, the area under the receiver operating characteristic curve (AUROC) was 96.6%. External testing performed across three studies showed AUROCs of 91.6-96.5%. In a separate test set of 186 eyes, we further compared the algorithm's performance with 4 ophthalmologists' evaluations. The algorithm performed comparably, if not being slightly more superior (sensitivity of 93.3% versus 51.7-96.6% by ophthalmologists and specificity of 99.0% versus 90.7-97.9% by ophthalmologists). Our findings show the potential of a retinal photograph-based screening tool for visually significant cataracts among older adults, providing more appropriate referrals to tertiary eye centers.

Deep Learning Automated Diagnosis and Quantitative Classification of Cataract Type and Severity

PURPOSE

To develop and evaluate deep learning models to perform automated diagnosis and quantitative classification of age-related cataract, including all three anatomical types, from anterior segment photographs.

DESIGN

Application of deep learning models to Age-Related Eye Disease Study (AREDS) dataset.

PARTICIPANTS

18,999 photographs (6,333 triplets) from longitudinal follow-up of 1,137 eyes (576 AREDS participants).

METHODS

Deep learning models were trained to detect and quantify nuclear cataract (NS; scale 0.9-7.1) from 45-degree slit-lamp photographs and cortical (CLO; scale 0-100%) and posterior subcapsular (PSC; scale 0-100%) cataract from retroillumination photographs. Model performance was compared with that of 14 ophthalmologists and 24 medical students. The ground truth labels were from reading center grading.

MAIN OUTCOME MEASURES

Mean squared error (MSE).

RESULTS

On the full test set, mean MSE values for the deep learning models were: 0.23 (SD 0.01) for NS, 13.1 (SD 1.6) for CLO, and 16.6 (SD 2.4) for PSC. On a subset of the test set (substantially enriched for positive cases of CLO and PSC), for NS, mean MSE for the models was 0.23 (SD 0.02), compared to 0.98 (SD 0.23; p=0.000001) for the ophthalmologists, and 1.24 (SD 0.33; p=0.000005) for the medical students. For CLO, mean MSE values were 53.5 (SD 14.8), compared to 134.9 (SD 89.9; p=0.003) and 422.0 (SD 944.4; p=0.0007), respectively. For PSC, mean MSE values were 171.9 (SD 38.9), compared to 176.8 (SD 98.0; p=0.67) and 395.2 (SD 632.5; p=0.18), respectively. In external validation on the Singapore Malay Eye Study (sampled to reflect the distribution of cataract severity in AREDS), MSE was 1.27 for NS and 25.5 for PSC.

CONCLUSIONS

A deep learning framework was able to perform automated and quantitative classification of cataract severity for all three types of age-related cataract. For the two most common types (NS and CLO), the accuracy was significantly superior to that of ophthalmologists; for the least common type (PSC), the accuracy was similar. The framework may have wide potential applications in both clinical and research domains. In the future, such approaches may increase the accessibility of cataract assessment globally. The code and models are publicly available at https://XXX.

Artificial Intelligence to Detect Meibomian Gland Dysfunction From in-vivo Laser Confocal Microscopy

Background: In recent years, deep learning has been widely used in a variety of ophthalmic diseases. As a common ophthalmic disease, meibomian gland dysfunction (MGD) has a unique phenotype in in-vivo laser confocal microscope imaging (VLCMI). The purpose of our study was to investigate a deep learning algorithm to differentiate and classify obstructive MGD (OMGD), atrophic MGD (AMGD) and normal groups. Methods: In this study, a multi-layer deep convolution neural network (CNN) was trained using VLCMI from OMGD, AMGD and healthy subjects as verified by medical experts. The automatic differential diagnosis of OMGD, AMGD and healthy people was tested by comparing its image-based identification of each group with the medical expert diagnosis. The CNN was trained and validated with 4,985 and 1,663 VLCMI images, respectively. By using established enhancement techniques, 1,663 untrained VLCMI images were tested. Results: In this study, we included 2,766 healthy control VLCMIs, 2,744 from OMGD and 2,801 from AMGD. Of the three models, differential diagnostic accuracy of the DenseNet169 CNN was highest at over 97%. The sensitivity and specificity of the DenseNet169 model for OMGD were 88.8 and 95.4%, respectively; and for AMGD 89.4 and 98.4%, respectively. Conclusion: This study described a deep learning algorithm to automatically check and classify VLCMI images of MGD. By optimizing the algorithm, the classifier model displayed excellent accuracy. With further development, this model may become an effective tool for the differential diagnosis of MGD.

Cataract: Advances in surgery and whether surgery remains the only treatment in future

Background

Cataract is the world's leading eye disease that causes blindness. The prevalence of cataract aged 40 years and older is approximately 11.8%-18.8%. Currently, surgery is the only way to treat cataracts.

Main Text

From early intracapsular cataract extraction to extracapsular cataract extraction, to current phacoemulsification cataract surgery, the incision ranges from 12 to 3 ​mm, and sometimes to even 1.8 ​mm or less, and the revolution in cataract surgery is ongoing. Cataract surgery has transformed from vision recovery to refractive surgery, leading to the era of refractive cataract surgery, and premium intraocular lenses (IOLs) such as toric IOLs, multifocal IOLs, and extended depth-of-focus IOLs are being increasingly used to meet the individual needs of patients. With its advantages of providing better visual acuity and causing fewer complications, phacoemulsification is currently the mainstream cataract surgery technique worldwide. However, patient expectations for the safety and accuracy of the operation are continually increasing. Femtosecond laser-assisted cataract surgery (FLACS) has entered the public's field of vision. FLACS is a combination of new laser technology and artificial intelligence to replace fine manual clear corneal incision, capsulorhexis, and nuclear pre-fragmentation, providing new alternative technologies for patients and ophthalmologists. As FLACS matures, it is being increasingly applied in complex cases; however, some think it is not cost-effective. Although more than 26 million cataract surgeries are performed each year, there is still a gap in the prevalence of cataracts, especially in developing countries. Although cataract surgery is a nearly ideal procedure and complications are manageable, both patients and doctors dream of using drugs to cure cataracts. Is surgery really the only way to treat cataracts in the future? It has been verified by animal experiments that lanosterol therapy in rabbits and dogs could make cataract severity alleviated and lens transparency partially recovered. Although there is still much to learn about cataract reversal, this groundbreaking work provided a new strategy for the prevention and treatment of cataracts.

Conclusions

Although cataract surgery is nearly ideal, it is still insufficient, we expect the prospects for cataract drugs to be bright.

A deep learning approach for successful big-bubble formation prediction in deep anterior lamellar keratoplasty

The efficacy of deep learning in predicting successful big-bubble (SBB) formation during deep anterior lamellar keratoplasty (DALK) was evaluated. Medical records of patients undergoing DALK at the University of Cologne, Germany between March 2013 and July 2019 were retrospectively analyzed. Patients were divided into two groups: (1) SBB or (2) failed big-bubble (FBB). Preoperative images of anterior segment optical coherence tomography and corneal biometric values (corneal thickness, corneal curvature, and densitometry) were evaluated. A deep neural network model, Visual Geometry Group-16, was selected to test the validation data, evaluate the model, create a heat map image, and calculate the area under the curve (AUC). This pilot study included 46 patients overall (11 women, 35 men). SBBs were more common in keratoconus eyes (KC eyes) than in corneal opacifications of other etiologies (non KC eyes) (p = 0.006). The AUC was 0.746 (95% confidence interval [CI] 0.603–0.889). The determination success rate was 78.3% (18/23 eyes) (95% CI 56.3–92.5%) for SBB and 69.6% (16/23 eyes) (95% CI 47.1–86.8%) for FBB. This automated system demonstrates the potential of SBB prediction in DALK. Although KC eyes had a higher SBB rate, no other specific findings were found in the corneal biometric data.

Deep-Learning-Based Pre-Diagnosis Assessment Module for Retinal Photographs: A Multicenter Study

Purpose

Artificial intelligence (AI) deep learning (DL) has been shown to have significant potential for eye disease detection and screening on retinal photographs in different clinical settings, particular in primary care. However, an automated pre-diagnosis image assessment is essential to streamline the application of the developed AI-DL algorithms. In this study, we developed and validated a DL-based pre-diagnosis assessment module for retinal photographs, targeting image quality (gradable vs. ungradable), field of view (macula-centered vs. optic-disc-centered), and laterality of the eye (right vs. left).

Methods

A total of 21,348 retinal photographs from 1914 subjects from various clinical settings in Hong Kong, Singapore, and the United Kingdom were used for training, internal validation, and external testing for the DL module, developed by two DL-based algorithms (EfficientNet-B0 and MobileNet-V2).

Results

For image-quality assessment, the pre-diagnosis module achieved area under the receiver operating characteristic curve (AUROC) values of 0.975, 0.999, and 0.987 in the internal validation dataset and the two external testing datasets, respectively. For field-of-view assessment, the module had an AUROC value of 1.000 in all of the datasets. For laterality-of-the-eye assessment, the module had AUROC values of 1.000, 0.999, and 0.985 in the internal validation dataset and the two external testing datasets, respectively.

Conclusions

Our study showed that this three-in-one DL module for assessing image quality, field of view, and laterality of the eye of retinal photographs achieved excellent performance and generalizability across different centers and ethnicities.

Translational Relevance

The proposed DL-based pre-diagnosis module realized accurate and automated assessments of image quality, field of view, and laterality of the eye of retinal photographs, which could be further integrated into AI-based models to improve operational flow for enhancing disease screening and diagnosis.