1. Artificial intelligence-enabled screening for diabetic retinopathy: a real-world, multicenter and prospective study
- Author
-
Qijuan Dong, Zilong Wang, Dong Zhao, Weiqing Wang, Juan Shi, Xun Xu, Jianjun Liu, Ying Peng, Yuancheng Dai, Yongde Peng, Zhiyun Zhao, Ling Hu, Heng Su, Fengmei Xu, Hongwei Jiang, Ziheng Zhou, Pei Gu, Kexin Qiu, Lei Chen, Yifei Zhang, Kun Liu, Qin Wan, Shengyin Jiao, Tingyu Ke, Xiaowei Ding, Demetri Terzopoulos, Li Yan, Benli Su, Guang Ning, and Qidong Zheng
- Subjects
Adult ,Research design ,China ,medicine.medical_specialty ,Endocrinology, Diabetes and Metabolism ,Concordance ,Fundus (eye) ,Diseases of the endocrine glands. Clinical endocrinology ,diagnostic techniques and procedures ,03 medical and health sciences ,chemistry.chemical_compound ,0302 clinical medicine ,Artificial Intelligence ,Diabetes mellitus ,Internal medicine ,Epidemiology ,Diabetes Mellitus ,medicine ,Humans ,Mass Screening ,Prospective Studies ,030212 general & internal medicine ,Prospective cohort study ,Aged ,Diabetic Retinopathy ,business.industry ,Diabetic retinopathy ,clinical study ,Middle Aged ,medicine.disease ,RC648-665 ,chemistry ,030221 ophthalmology & optometry ,Epidemiology/Health services research ,epidemiology ,Female ,Glycated hemoglobin ,business - Abstract
IntroductionEarly screening for diabetic retinopathy (DR) with an efficient and scalable method is highly needed to reduce blindness, due to the growing epidemic of diabetes. The aim of the study was to validate an artificial intelligence-enabled DR screening and to investigate the prevalence of DR in adult patients with diabetes in China.Research design and methodsThe study was prospectively conducted at 155 diabetes centers in China. A non-mydriatic, macula-centered fundus photograph per eye was collected and graded through a deep learning (DL)-based, five-stage DR classification. Images from a randomly selected one-third of participants were used for the DL algorithm validation.ResultsIn total, 47 269 patients (mean (SD) age, 54.29 (11.60) years) were enrolled. 15 805 randomly selected participants were reviewed by a panel of specialists for DL algorithm validation. The DR grading algorithms had a 83.3% (95% CI: 81.9% to 84.6%) sensitivity and a 92.5% (95% CI: 92.1% to 92.9%) specificity to detect referable DR. The five-stage DR classification performance (concordance: 83.0%) is comparable to the interobserver variability of specialists (concordance: 84.3%). The estimated prevalence in patients with diabetes detected by DL algorithm for any DR, referable DR and vision-threatening DR were 28.8% (95% CI: 28.4% to 29.3%), 24.4% (95% CI: 24.0% to 24.8%) and 10.8% (95% CI: 10.5% to 11.1%), respectively. The prevalence was higher in female, elderly, longer diabetes duration and higher glycated hemoglobin groups.ConclusionThis study performed, a nationwide, multicenter, DL-based DR screening and the results indicated the importance and feasibility of DR screening in clinical practice with this system deployed at diabetes centers.Trial registration numberNCT04240652.
- Published
- 2020