Abstract

Screening for chronic kidney disease is a challenge in community and primary care settings, even in high-income countries. We developed an artificial intelligence deep learning algorithm (DLA) to detect chronic kidney disease from retinal images, which could add to existing chronic kidney disease screening strategies. We used data from three population-based, multiethnic, cross-sectional studies in Singapore and China. The Singapore Epidemiology of Eye Diseases study (SEED, patients aged ≥40 years) was used to develop (5188 patients) and validate (1297 patients) the DLA. External testing was done on two independent datasets: the Singapore Prospective Study Program (SP2, 3735 patients aged ≥25 years) and the Beijing Eye Study (BES, 1538 patients aged ≥40 years). Chronic kidney disease was defined as estimated glomerular filtration rate less than 60 mL/min per 1·73m2. Three models were trained: 1) image DLA; 2) risk factors (RF) including age, sex, ethnicity, diabetes, and hypertension; and 3) hybrid DLA combining image and RF. Model performances were evaluated using the area under the receiver operating characteristic curve (AUC). In the SEED validation dataset, the AUC was 0·911 for image DLA (95% CI 0·886 -0·936), 0·916 for RF (0·891-0·941), and 0·938 for hybrid DLA (0·917-0·959). Corresponding estimates in the SP2 testing dataset were 0·733 for image DLA (95% CI 0·696-0·770), 0·829 for RF (0·797-0·861), and 0·810 for hybrid DLA (0·776-0·844); and in the BES testing dataset estimates were 0·835 for image DLA (0·767-0·903), 0·887 for RF (0·828-0·946), and 0·858 for hybrid DLA (0·794-0·922). AUC estimates were similar in subgroups of people with diabetes (image DLA 0·889 [95% CI 0·850-0·928], RF 0·899 [0·862-0·936], hybrid 0·925 [0·893-0·957]) and hypertension (image DLA 0·889 [95% CI 0·860-0·918], RF 0·889 [0·860-0·918], hybrid 0·918 [0·893-0·943]). A retinal image DLA shows good performance for estimating chronic kidney disease, underlying the feasibility of using retinal photography as an adjunctive or opportunistic screening tool for chronic kidney disease in community populations. National Medical Research Council, Singapore.

Highlights

  • Chronic kidney disease is a major global public health problem.[1]

  • area under the curve (AUC) of the deep learning algorithm (DLA) for image-only, risk factor (RF), and the hybrid model are shown in figure 2

  • AUC was higher than image-only DLA in both SP2 (0·810 vs 0·733; p=0·0005) and Beijing Eye Study (BES) (0·858 vs 0·835; p=0·0002)

Read more

Summary

Introduction

Chronic kidney disease is a major global public health problem.[1] Because earlier detection of chronic kidney disease allows appropriate interventions, regular screening is recommended for the general population,[2,3] and in high-risk populations (eg, patients with diabetes or hypertension, and specific ethnic groups).[4] screening of chronic kidney disease depends on measurement of the estimated glomerular filtration rate (eGFR, calculated from serum creatinine), or urine tests for protein or albumin. Because serum or urine samples must be obtained, adherence to screening is low, even in high-income countries and in at-risk populations. A study in Australia[5] showed that nearly 50% of patients with diabetes attending general practice had not been screened for chronic kidney disease in the previous 18 months.[5] a urine sample is less invasive and easier to obtain, albuminuria is highly variable, with intra-individual variation of up to 50%.6

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call