Carol Y. Cheung
A5019493583- Retinal Imaging and Analysis
- Retinal Diseases and Treatments
- Glaucoma and retinal disorders
- Retinal and Optic Conditions
- Optical Coherence Tomography Applications
- Ophthalmology and Visual Impairment Studies
- Corneal surgery and disorders
- Digital Imaging for Blood Diseases
- Blood Pressure and Hypertension Studies
- Ophthalmology and Visual Health Research
- Dementia and Cognitive Impairment Research
- Cerebral Venous Sinus Thrombosis
- Artificial Intelligence in Healthcare
- Artificial Intelligence in Healthcare and Education
- Neonatal and fetal brain pathology
- Ocular Surface and Contact Lens
- Retinal and Macular Surgery
- Preterm Birth and Chorioamnionitis
- Intraocular Surgery and Lenses
- Acute Ischemic Stroke Management
- Retinopathy of Prematurity Studies
- Autopsy Techniques and Outcomes
- Ergonomics and Musculoskeletal Disorders
- Ocular and Laser Science Research
- COVID-19 diagnosis using AI
Chinese University of Hong Kong
2017-2025
Hong Kong Eye Hospital
2018-2025
Sun Yat-sen University
2023
Creative Commons
2021
Singapore National Eye Center
2012-2020
Singapore Eye Research Institute
2012-2020
The University of Sydney
2012-2019
Duke-NUS Medical School
2012-2018
University of Hong Kong
2018
National University of Singapore
2012-2018
Nonophthalmologist physicians do not confidently perform direct ophthalmoscopy. The use of artificial intelligence to detect papilledema and other optic-disk abnormalities from fundus photographs has been well studied.We trained, validated, externally tested a deep-learning system classify optic disks as being normal or having 15,846 retrospectively collected ocular that had obtained with pharmacologic pupillary dilation various digital cameras in persons multiple ethnic populations. Of...
BackgroundScreening for chronic kidney disease is a challenge in community and primary care settings, even high-income countries. We developed an artificial intelligence deep learning algorithm (DLA) to detect from retinal images, which could add existing screening strategies.MethodsWe used data three population-based, multiethnic, cross-sectional studies Singapore China. The Epidemiology of Eye Diseases study (SEED, patients aged ≥40 years) was develop (5188 patients) validate (1297 the...
Abstract Primary diabetes care and diabetic retinopathy (DR) screening persist as major public health challenges due to a shortage of trained primary physicians (PCPs), particularly in low-resource settings. Here, bridge the gaps, we developed an integrated image–language system (DeepDR-LLM), combining large language model (LLM module) image-based deep learning (DeepDR-Transformer), provide individualized management recommendations PCPs. In retrospective evaluation, LLM module demonstrated...
This study explores the potential of Artificial Intelligence (AI) in early screening and prognosis Dry Eye Disease (DED), aiming to enhance accuracy therapeutic approaches for eye-care practitioners. Despite promising opportunities, challenges such as diverse diagnostic evidence, complex etiology, interdisciplinary knowledge integration impede interpretability, reliability, applicability AI-based DED detection methods. The research conducts a comprehensive review datasets, standards, well...
BackgroundSpectral-domain optical coherence tomography (SDOCT) can be used to detect glaucomatous optic neuropathy, but human expertise in interpretation of SDOCT is limited. We aimed develop and validate a three-dimensional (3D) deep-learning system using volumes neuropathy.MethodsWe retrospectively collected dataset including 4877 disc cube for training (60%), testing (20%), primary validation (20%) from electronic medical research records at the Chinese University Hong Kong Eye Centre...
To determine the myopia prevalence in Hong Kong Chinese children and their parents.It was a population-based cross-sectional study. A total of 4257 aged 6-8 years, 5880 parents were recruited Children Eye Study. Cycloplegic autorefraction measured for children; non-cycloplegic parents. Parental educational level, children's outdoor time, near work collected by validated questionnaires.In 25.0% myopic, among them, 12.7% 6-year-olds, 24.4% 7-year-olds 36.1% 8-year-old. About 0.7% 8 years high...
<h2>Summary</h2><h3>Background</h3> Ocular changes are traditionally associated with only a few hepatobiliary diseases. These non-specific and have low detection rate, limiting their potential use as clinically independent diagnostic features. Therefore, we aimed to engineer deep learning models establish associations between ocular features major diseases advance automated screening identification of from images. <h3>Methods</h3> We did multicentre, prospective study develop using slit-lamp...
To validate a vascular severity score as an appropriate output for artificial intelligence (AI) Software Medical Device (SaMD) retinopathy of prematurity (ROP) through comparison with ordinal disease labels stage and plus assigned by the International Classification Retinopathy Prematurity, Third Edition (ICROP3), committee.Validation study AI-based ROP score.A total 34 experts from ICROP3 committee.Two separate datasets 30 fundus photographs each (0-5) (plus, preplus, neither) were labeled...
To assess the diagnostic performance of two chatbots, ChatGPT and Glass, in uveitis diagnosis compared to renowned specialists, evaluate clinicians' perception about utilizing artificial intelligence (AI) ophthalmology practice.
The increasing prevalence of myopia worldwide presents a significant public health challenge. A key strategy to combat is with early detection and prediction in children as such examination allows for effective intervention using readily accessible imaging technique. To this end, we introduced DeepMyopia, an artificial intelligence (AI)-enabled decision support system detect predict onset facilitate targeted interventions at risk routine retinal fundus images. Based on deep learning...
In any community, the key to understanding burden of a specific condition is conduct an epidemiological study. The deep learning system (DLS) recently showed promising diagnostic performance for diabetic retinopathy (DR). This study aims use DLS as grading tool, instead human assessors, determine prevalence and systemic cardiovascular risk factors DR on fundus photographs, in patients with diabetes. multi-ethnic (5 races), multi-site (8 datasets from Singapore, USA, Hong Kong, China...
Systematic or national screening programs for diabetic retinopathy (DR) and macular edema (DME), using digital fundus photography optical coherence tomography (OCT), are currently implemented at primary care level, aiming to provide timely referral vision-threatening DR DME ophthalmologists treatment vision loss prevention. However, interpretation of retinal images requires specialized knowledge expertise in eye disease. Furthermore, current capital- labor-intensive, which makes it difficult...
Aims To evaluate the distributions of quantitative optical coherence tomography angiography (OCT-A) metrics and its associated factors in children. Methods 1059 children aged 6–8 years were recruited from Hong Kong Children Eye Study. All participants underwent OCT-A with a swept-source OCT. Retinal microvasculature on superficial capillary plexus was assessed quantified by customised automated image analysis programme. Univariable multiple linear regression analyses performed to determine...
Deep learning has achieved remarkable success in the optical coherence tomography (OCT) image classification task with substantial labelled B-scan images available. However, obtaining such fine-grained expert annotations is usually quite difficult and expensive. How to leverage volume-level labels develop a robust classifier very appealing. In this paper, we propose weakly supervised deep framework uncertainty estimation address macula-related disease problem from OCT only label being First,...