Ching-yu Cheng

• MD, MPH, PhD
• Professor at National University of Singapore

Background To investigate the relationship between obesity and corneal nerve metrics in patients with type 2 diabetes mellitus (DM). Methods This cross-sectional study included a total of 385 healthy controls and 663 patients with DM. Metrics for corneal nerve and epithelial cells were evaluated using in-vivo confocal microscopy (IVCM). Corneal ne...

Artificial intelligence (AI) shows remarkable potential in medical imaging diagnostics, but current models typically require retraining when deployed across different clinical centers, limiting their widespread adoption. We introduce GlobeReady, a clinician-friendly AI platform that enables ocular disease diagnosis without retraining/fine-tuning or...

Background The advent of generative artificial intelligence has led to the emergence of multiple vision large language models (VLLMs). This study aimed to evaluate the capabilities of commonly available VLLMs, such as OpenAI’s GPT-4V and Google’s Gemini, in detecting and diagnosing ocular diseases from retinal images. Methods and analysis From the...

Rationale: Chronic kidney disease (CKD) is a major public health problem worldwide associated with cardiovascular disease, renal failure, and mortality. To effectively address this growing burden, innovative solutions to management are urgently required. We conducted a scoping review to identify key use cases in which artificial intelligence (AI) c...

Precision medicine (PM) research in recent years has witnessed a remarkable surge in large-scale population genomics programs. In 2017, Singapore initiated the National Precision Medicine (NPM) program, a three-phase national strategy driving PM research, innovation, and enterprise capitalizing on Singapore’s unique multi-ancestral Asian population...

INTRODUCTION The utility of retinal photography‐derived aging biomarkers for predicting cognitive decline remains under‐explored. METHODS A memory‐clinic cohort in Singapore was followed‐up for 5 years. RetiPhenoAge, a retinal aging biomarker, was derived from retinal photographs using deep‐learning. Using competing risk analysis, we determined th...

The eye provides novel insights into general health, as well as pathogenesis and development of systemic diseases. In the past decade, growing evidence has demonstrated that the eye's structure and function mirror multiple systemic health conditions, especially in cardiovascular diseases, neurodegenerative disorders, and kidney impairments. This ha...

Background Previously, based on retinal photographs, we developed a deep‐learning algorithm to predict biological age (termed, RetiAGE) that was associated with future risks of morbidity and mortality. This study specifically aimed to evaluate the performance of RetiAGE in predicting future risks of chronic obstructive pulmonary disease (COPD). Me...

Brain age has emerged as a powerful tool to understand neuroanatomical aging and its link to health outcomes like cognition. However, there remains a lack of studies investigating the rate of brain aging and its relationship to cognition. Furthermore, most brain age models are trained and tested on cross-sectional data from primarily Caucasian, adu...

Ocular diseases, including diabetic retinopathy and glaucoma, present a significant public health challenge due to their high prevalence and potential for causing vision impairment. Early and accurate diagnosis is crucial for effective treatment and management. In recent years, deep learning models have emerged as powerful tools for analysing medic...

Diabetes poses a considerable global health challenge, with varying levels of diabetes knowledge among healthcare professionals, highlighting the importance of diabetes training. Large Language Models (LLMs) provide new insights into diabetes training, but their performance in diabetes-related queries remains uncertain, especially outside the Engli...

Background: RETFound, a self-supervised, retina-specific foundation model (FM), showed potential in downstream applications. However, its comparative performance with traditional deep learning (DL) models remains incompletely understood. This study aimed to evaluate RETFound against three ImageNet-pretrained supervised DL models (ResNet50, ViT-base...

Background To estimate the additive associations of cardiometabolic multimorbidity (CMM) and depression on long‐term cognitive trajectory in multi‐regional cohorts and validate the generalizability of the findings in varying clinical settings. Method Data harmonization was performed across 14 longitudinal cohort studies within the Cohort Studies o...

Background: To evaluate the 6-year physiological rates-of-change in ganglion cell inner plexiform layer (GCIPL) and retinal nerve fibre layer (RNFL) thickness measured with optical coherence tomography. Methods: We included 2202 out of 2661 subjects from the population-based Singapore Chinese Eye Study who returned for follow-up 6 years after ba...

Objectives: To determine the association between telomere length (TL) and age-related macular degeneration (AMD) and examine the potential variations with sex and ethnicity. Methods: Population-based, cross-sectional study. A total of 52,083 participants from the UK Biobank were included. Leucocyte TL, measured using quantitative polymerase chai...

BACKGROUND The American Heart Association recently published guidelines on how to clinically identify and categorize individuals with cardiovascular-kidney-metabolic (CKM) syndrome. The extent to which CKM syndrome prevalence and prognosis differ by sex remains unknown. This study aimed to examine the impact of sex on trends in prevalence over 30 y...

Purpose: To estimate the prevalence of vision loss for 2020 in South and Central Asia and analyze trends since 1990. Methods: In a systematic literature review, we estimated the prevalence of blindness, visual impairment (VI) and presbyopia-related VI in 1990,2000,2010, and 2020. Results: The study included 103 population-based studies. In South/...

Purpose To elucidate the genetic basis of primary angle-closure glaucoma (PACG) by identifying pathogenic tissue and critical tissue-specific variants. Methods The correlations among PACG susceptibility, axial length (AL), and anterior chamber depth (ACD) were evaluated using meta-analyses. Propensity score matching was utilized on 2161 participan...

Inability to express the confidence level and detect unseen disease classes limits the clinical implementation of artificial intelligence in the real world. We develop a foundation model with uncertainty estimation (FMUE) to detect 16 retinal conditions on optical coherence tomography (OCT). In the internal test set, FMUE achieves a higher F1 score...

Purpose We investigated the association between metabolically healthy obesity (MHO) and retinal age gap and explored potential sex differences in this association. Methods This study included 30,335 participants from the UK Biobank. Body mass index (BMI) was classified into normal weight, overweight, and obesity. Metabolic health (MH) was defined...

Primary open-angle glaucoma typically presents as two subtypes. This study aimed to elucidate the shared and distinct genetic architectures of normal-tension (NTG) and high-tension glaucoma (HTG), motivated by the need to develop intraocular pressure (IOP)-independent drug targets for the disease. We conducted a comprehensive multi-ethnic meta-anal...

Background: The American Heart Association recently published guidelines on how to clinically identify and categorize individuals with cardiovascular-kidney-metabolic (CKM) syndrome. The extent to which CKM syndrome prevalence and prognosis differ by sex remain unknown. This study aimed to examine the impact of sex on trends in prevalence over 30 y...

Introduction: Reti-CVD, a novel cardiovascular risk stratification tool derived from deep learning and retinal photography, offers a promising alternative to CT scan-measured Coronary Artery Calcium (CAC) scores in predicting cardiovascular events. This study investigates the profile of Reti-CVD scores using data from three medical centers in South...

Introduction: Cardiovascular diseases (CVD) are the leading cause of death in developed countries. Coronary artery calcium (CAC) is a clinically validated strong marker of CVD, and previous studies suggest that retinal blood vessels provide relevant information. This study aimed to validate the Reti-CVD model, developed for predicting CAC score thr...

Structural variants (SVs) are significant contributors to inter-individual genetic variation associated with traits and diseases. Current SV studies using whole-genome sequencing (WGS) have a largely Eurocentric composition, with little known about SV diversity in other ancestries, particularly from Asia. Here, we present a WGS catalogue of 73,035...

Background/ Aims The lack of context for anterior segment optical coherence tomography (ASOCT) measurements impedes its clinical utility. We established the normative distribution of anterior chamber depth (ACD), area (ACA) and width (ACW) and lens vault (LV), and applied percentile cut-offs to detect primary angle closure disease (PACD; ≥180° post...

Cardiovascular disease (CVD) is the leading cause of death in Asians. We aimed to examine the validity and reliability of self-reported (SR) CVD in 6762 Chinese, Malay, and Indian adults aged 40–80 years who attended the baseline (November 2004) and 6-year follow-up visit (2011–2017) of a population-based cohort study in Singapore. CVD was defined...

We investigated whether the effect of lipid-lowering drugs (LLDs) on age-related macular degeneration (AMD) differs according to the main complement genetic variants in Singapore Epidemiology of Eye Diseases (SEED) (n = 5,579) and UK Biobank studies (n = 445,727). The effect of LLD was determined for each stratum of 20 complement genetic variants....

Purpose Recent studies utilized ocular images and deep learning (DL) to predict refractive error and yielded notable results. However, most studies did not address biases from imbalanced datasets or conduct external validations. To address these gaps, this study aimed to integrate the deep imbalanced regression (DIR) technique into ResNet and Visio...

Community vision screening plays a crucial role in identifying individuals with vision loss and preventing avoidable blindness, particularly in rural communities where access to eye care services is limited. Currently, there is a pressing need for a simple and efficient process to screen and refer individuals with significant eye disease-related vi...

Objective Our objectives were to identify correlation patterns between complement and lipid pathways using a multiomics data integration approach and to determine how these interconnections affect age-related macular degeneration (AMD). Design Nested case-control study. Subjects and Controls The analyses were performed in a subset of the Singapor...

Importance Myopic maculopathy (MM) is a major cause of vision impairment globally. Artificial intelligence (AI) and deep learning (DL) algorithms for detecting MM from fundus images could potentially improve diagnosis and assist screening in a variety of health care settings. Objectives To evaluate DL algorithms for MM classification and segmentat...

Timely disease diagnosis is challenging due to increasing disease burdens and limited clinician availability. AI shows promise in diagnosis accuracy but faces real-world application issues due to insufficient validation in clinical workflows and diverse populations. This study addresses gaps in medical AI downstream accountability through a case st...

Background Cardiometabolic multimorbidity (CMM) and depression are often co-occurring in older adults and associated with neurodegenerative outcomes. The present study aimed to estimate the independent and joint associations of CMM and depression on cognitive function in multi-regional cohorts, and to validate the generalizability of the findings i...

Recent advancements in artificial intelligence (AI), particularly in deep learning and large language models (LLMs), have accelerated their integration into medicine. However, these developments have also raised public concerns about the safe application of AI. In healthcare, these concerns are especially pertinent, as the ethical and secure deploy...

The study aimed to evaluate the impact of compensating retinal nerve fiber layer (RNFL) thickness for demographic and anatomical factors on glaucoma detection in Chinese and Indian adults. A population‐based study included 1995 healthy participants (1076 Chinese and 919 Indians) to construct a multivariable linear regression compensation model. Thi...

Background Housing has been associated with dementia risk and disability, but associations of housing with differential patterns of neuropsychiatric symptoms (NPS) among dementia-free older adults remain to be explored. The present study sought to explore the contribution of housing status on NPS and subsyndromes associated with cognitive dysfuncti...

Retinal foundation models aim to learn generalizable representations from diverse retinal images, facilitating label-efficient model adaptation across various ophthalmic tasks. Despite their success, current retinal foundation models are generally restricted to a single imaging modality, such as Color Fundus Photography (CFP) or Optical Coherence T...

Purpose To predict 10-year graft survival after deep anterior lamellar keratoplasty (DALK) and penetrating keratoplasty (PK) using a machine learning (ML)-based interpretable risk score. Methods Singapore Corneal Transplant Registry patients (n = 1687) who underwent DALK (n = 524) or PK (n = 1163) for optical indications (excluding endothelial dis...

Aims To develop and externally test deep learning (DL) models for assessing the image quality of three-dimensional (3D) macular scans from Cirrus and Spectralis optical coherence tomography devices. Methods We retrospectively collected two data sets including 2277 Cirrus 3D scans and 1557 Spectralis 3D scans, respectively, for training (70%), fine...

Primary diabetes care and diabetic retinopathy (DR) screening persist as major public health challenges due to a shortage of trained primary care physicians (PCPs), particularly in low-resource settings. Here, to bridge the gaps, we developed an integrated image–language system (DeepDR-LLM), combining a large language model (LLM module) and image-b...

The Meishan-Chiayi area of western Taiwan has a large probability of producing a major earthquake in the near future. Historically, one of the largest and most damaging of Taiwan’s earthquakes occurred there. It is, therefore, important to have a well-constrained upper crustal 3-D shear-wave velocity model that can be used to accurately determine g...

Background To estimate global and regional trends from 2000 to 2020 of the number of persons visually impaired by cataract and their proportion of the total number of vision-impaired individuals. Methods A systematic review and meta-analysis of published population studies and gray literature from 2000 to 2020 was carried out to estimate global a...

Objective The aim of this study was to evaluate the accuracy, comprehensiveness, and safety of a publicly available large language model (LLM)—ChatGPT in the sub-domain of glaucoma. Design Evaluation of diagnostic test or technology. Subjects, participants, and/or controls We seek to evaluate the responses of an artificial intelligence chatbot Ch...

Background We aimed to update estimates of global vision loss due to age-related macular degeneration (AMD). Methods We did a systematic review and meta-analysis of population-based surveys of eye diseases from January, 1980, to October, 2018. We fitted hierarchical models to estimate the prevalence of moderate and severe vision impairment (MSVI;...

Background Uncorrected refractive error is a major cause of vision impairment worldwide and its increasing prevalent necessitates effective screening and management strategies. Meanwhile, deep learning, a subset of Artificial Intelligence, has significantly advanced ophthalmological diagnostics by automating tasks that required extensive clinical e...

Objectives To estimate global and regional trends from 2000 to 2020 of the number of persons visually impaired by diabetic retinopathy and their proportion of the total number of vision-impaired individuals. Methods Data from population-based studies on eye diseases between 1980 to 2018 were compiled. Meta-regression models were performed to estim...

The current retinal artificial intelligence models were trained using data with a limited category of diseases and limited knowledge. In this paper, we present a retinal vision-language foundation model (RetiZero) with knowledge of over 400 fundus diseases. Specifically, we collected 341,896 fundus images paired with text descriptions from 29 publi...

Hypertensive retinopathy (HR) can potentially lead to vision loss if left untreated. Early screening and treatment are critical in reducing the risk of vision loss. The computer-aided diagnostic system presents an opportunity to improve the efficiency and reliability of HR screening and diagnosis, particularly given the shortage of specialized medi...

Brain age has emerged as a powerful tool to understand neuroanatomical aging and its link to health outcomes like cognition. However, there remains a lack of studies investigating the rate of brain aging and its relationship to cognition. Furthermore, most brain age models are trained and tested on cross-sectional data from primarily Caucasian, adu...

Brain age has emerged as a powerful tool to understand neuroanatomical aging and its link to health outcomes like cognition. However, there remains a lack of studies investigating the rate of brain aging and its relationship to cognition. Furthermore, most brain age models are trained and tested on cross-sectional data from primarily Caucasian, adu...

Background Artificial intelligence (AI) that utilizes deep learning (DL) has potential for systemic disease prediction using retinal imaging. The retina’s unique features enable non-invasive visualization of the central nervous system and microvascular circulation, aiding early detection and personalized treatment plans for personalized care. This...

The size of the human head is highly heritable, but genetic drivers of its variation within the general population remain unmapped. We perform a genome-wide association study on head size (N = 80,890) and identify 67 genetic loci, of which 50 are novel. Neuroimaging studies show that 17 variants affect specific brain areas, but most have widespread...

Objective Vision transformers (ViTs) have shown promising performance in various classification tasks previously dominated by convolutional neural networks (CNNs). However, the performance of ViTs in referable diabetic retinopathy (DR) detection is relatively underexplored. In this study, using retinal photographs, we evaluated the comparative perf...

Background To determine the prevalence, risk factors; and impact on patient health and economic outcomes across the laterality spectrum of multiple sensory impairment (MSI) in a multi-ethnic older Asian population. Methods In this population-based study of Singaporeans aged ≥ 60 years, MSI was defined as concomitant vision (visual acuity > 0.3 log...

Objective Our objective was to determine the effects of lipids and complement proteins on early and intermediate age-related macular degeneration (AMD) stages using machine learning models by integrating metabolomics and proteomic data. Design Nested case–control study. Subjects and Controls The analyses were performed in a subset of the Singapor...

Background Diabetic kidney disease (DKD) and diabetic retinopathy (DR) are major diabetic microvascular complications, contributing significantly to morbidity, disability, and mortality worldwide. The kidney and the eye, having similar microvascular structures and physiological and pathogenic features, may experience similar metabolic changes in di...

Introduction The clinical presentations of dry eye disease (DED) and depression (DEP) often comanifest. However, the robustness and the mechanisms underlying this association were undetermined. Objectives To this end, we set up a three-segment study that employed multimodality results (meta-analysis, genome-wide association study [GWAS] and Mendel...

Background: To estimate global and regional trends from 2000 to 2020 of the number of persons visually impaired by cataract and their proportion of the total number of vision-impaired individuals. Methods: A systematic review and meta-analysis of published population studies and gray literature from 2000 to 2020 was carried out to estimate global...

Type 2 diabetes (T2D) is a heterogeneous disease that develops through diverse pathophysiological processes1,2 and molecular mechanisms that are often specific to cell type3,4. Here, to characterize the genetic contribution to these processes across ancestry groups, we aggregate genome-wide association study data from 2,535,601 individuals (39.7% n...

The prevalence of chronic kidney disease (CKD) is high. Identification of cases with CKD or at high risk of developing it is important to tailor early interventions. The objective of this study was to identify blood metabolites associated with prevalent and incident severe CKD, and to quantify the corresponding improvement in CKD detection and pred...

X-chromosomal genetic variants are understudied but can yield valuable insights into sexually dimorphic human traits and diseases. We performed a sex-stratified cross-ancestry X-chromosome-wide association meta-analysis of seven kidney-related traits (n = 908,697), identifying 23 loci genome-wide significantly associated with two of the traits: 7 f...

Primary open-angle glaucoma (POAG), characterized by retinal ganglion cell death, is a leading cause of irreversible blindness worldwide. However, its molecular and cellular causes are not well understood. Elevated intraocular pressure (IOP) is a major risk factor, but many patients have normal IOP. Colocalization and Mendelian randomization analys...

Diabetic retinopathy (DR) is the leading cause of preventable blindness worldwide. The risk of DR progression is highly variable among different individuals, making it difficult to predict risk and personalize screening intervals. We developed and validated a deep learning system (DeepDR Plus) to predict time to DR progression within 5 years solely...

Background To investigate the discriminant validity and time‐effectiveness of a stepwise dementia case‐finding approach in a community‐based Singaporean older adult population. Method Participants who completed the Progressive Forgetfulness Question (PFQ) and the Abbreviated Mental Test (AMT) were invited to phase II and administered the Mini‐Ment...

Background Cataract diagnosis typically requires in-person evaluation by an ophthalmologist. However, color fundus photography (CFP) is widely performed outside ophthalmology clinics, which could be exploited to increase the accessibility of cataract screening by automated detection. Methods DeepOpacityNet was developed to detect cataracts from CF...

Brain age has emerged as a powerful tool to understand neuroanatomical aging and its link to health outcomes like cognition. However, most brain age models are trained and tested on cross-sectional data from primarily Caucasian, adult participants. It is thus unclear how well these models generalize to non-Caucasian participants, especially childre...

Background Retinal photographs allow a non-invasive way to see the human vasculature and provide insights into cardiovascular disease (CVD). In our previous study, we developed the Reti-CVD, a deep-learning algorithm to predict the future CVD events from retinal photographs. Purpose In this study, we extend the application of Reti-CVD by investiga...

Introduction: The advent of sophisticated deep learning algorithms has now made it possible to predict the risk of cardiovascular diseases (CVDs) using retinal images. We had previously developed a retina-based deep learning model, Reti-CVD, trained on coronary artery calcium (CAC) data, which successfully predicted future CVD incidents in a longit...

Introduction: This study aimed to assess the ability of a deep learning algorithm, Reti-AF, developed from retinal photos, to predict atrial fibrillation (AF) incidence. Its predictive performance was evaluated using the UK Biobank and compared to Reti-CAC, a similar algorithm trained on coronary artery calcium (CAC) scores. Hypothesis: We hypothes...

Introduction: Our previous work led to developing a deep learning algorithm for retinal images, Reti-CVD, which effectively predicted cardiovascular disease (CVD) events in individuals without CVD history, leveraging coronary artery calcium (CAC) scores for algorithm training. Hypothesis: This study aims to assess the capability of deep learning-as...

Introduction: Educational attainment, widely used in epidemiologic studies as a surrogate for socioeconomic status, is a predictor of cardiovascular health outcomes. Methods: A two-stage genome-wide meta-analysis of low-density lipoprotein cholesterol (LDL), high-density lipoprotein cholesterol (HDL), and triglyceride (TG) levels was performed whil...