Advancing healthcare with artificial intelligence: diagnostic accuracy of machine learning algorithm in diagnosis of diabetic retinopathy in the Brazilian population.
Mateus A Dos Reis, Cristiano A Künas, Thiago da Silva Araújo, Josiane Schneiders, Pietro B de Azevedo, Luis F Nakayama, Dimitris R V Rados, Roberto N Umpierre, Otávio Berwanger, Daniel Lavinsky, Fernando K Malerbi, Philippe O A Navaux, Beatriz D Schaan
{"title":"Advancing healthcare with artificial intelligence: diagnostic accuracy of machine learning algorithm in diagnosis of diabetic retinopathy in the Brazilian population.","authors":"Mateus A Dos Reis, Cristiano A Künas, Thiago da Silva Araújo, Josiane Schneiders, Pietro B de Azevedo, Luis F Nakayama, Dimitris R V Rados, Roberto N Umpierre, Otávio Berwanger, Daniel Lavinsky, Fernando K Malerbi, Philippe O A Navaux, Beatriz D Schaan","doi":"10.1186/s13098-024-01447-0","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>In healthcare systems in general, access to diabetic retinopathy (DR) screening is limited. Artificial intelligence has the potential to increase care delivery. Therefore, we trained and evaluated the diagnostic accuracy of a machine learning algorithm for automated detection of DR.</p><p><strong>Methods: </strong>We included color fundus photographs from individuals from 4 databases (primary and specialized care settings), excluding uninterpretable images. The datasets consist of images from Brazilian patients, which differs from previous work. This modification allows for a more tailored application of the model to Brazilian patients, ensuring that the nuances and characteristics of this specific population are adequately captured. The sample was fractionated in training (70%) and testing (30%) samples. A convolutional neural network was trained for image classification. The reference test was the combined decision from three ophthalmologists. The sensitivity, specificity, and area under the ROC curve of the algorithm for detecting referable DR (moderate non-proliferative DR; severe non-proliferative DR; proliferative DR and/or clinically significant macular edema) were estimated.</p><p><strong>Results: </strong>A total of 15,816 images (4590 patients) were included. The overall prevalence of any degree of DR was 26.5%. Compared with human evaluators (manual method of diagnosing DR performed by an ophthalmologist), the deep learning algorithm achieved an area under the ROC curve of 0.98 (95% CI 0.97-0.98), with a specificity of 94.6% (95% CI 93.8-95.3) and a sensitivity of 93.5% (95% CI 92.2-94.9) at the point of greatest efficiency to detect referable DR.</p><p><strong>Conclusions: </strong>A large database showed that this deep learning algorithm was accurate in detecting referable DR. This finding aids to universal healthcare systems like Brazil, optimizing screening processes and can serve as a tool for improving DR screening, making it more agile and expanding care access.</p>","PeriodicalId":11106,"journal":{"name":"Diabetology & Metabolic Syndrome","volume":null,"pages":null},"PeriodicalIF":3.4000,"publicationDate":"2024-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11360296/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Diabetology & Metabolic Syndrome","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1186/s13098-024-01447-0","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENDOCRINOLOGY & METABOLISM","Score":null,"Total":0}
引用次数: 0
Abstract
Background: In healthcare systems in general, access to diabetic retinopathy (DR) screening is limited. Artificial intelligence has the potential to increase care delivery. Therefore, we trained and evaluated the diagnostic accuracy of a machine learning algorithm for automated detection of DR.
Methods: We included color fundus photographs from individuals from 4 databases (primary and specialized care settings), excluding uninterpretable images. The datasets consist of images from Brazilian patients, which differs from previous work. This modification allows for a more tailored application of the model to Brazilian patients, ensuring that the nuances and characteristics of this specific population are adequately captured. The sample was fractionated in training (70%) and testing (30%) samples. A convolutional neural network was trained for image classification. The reference test was the combined decision from three ophthalmologists. The sensitivity, specificity, and area under the ROC curve of the algorithm for detecting referable DR (moderate non-proliferative DR; severe non-proliferative DR; proliferative DR and/or clinically significant macular edema) were estimated.
Results: A total of 15,816 images (4590 patients) were included. The overall prevalence of any degree of DR was 26.5%. Compared with human evaluators (manual method of diagnosing DR performed by an ophthalmologist), the deep learning algorithm achieved an area under the ROC curve of 0.98 (95% CI 0.97-0.98), with a specificity of 94.6% (95% CI 93.8-95.3) and a sensitivity of 93.5% (95% CI 92.2-94.9) at the point of greatest efficiency to detect referable DR.
Conclusions: A large database showed that this deep learning algorithm was accurate in detecting referable DR. This finding aids to universal healthcare systems like Brazil, optimizing screening processes and can serve as a tool for improving DR screening, making it more agile and expanding care access.
背景：在一般的医疗保健系统中，糖尿病视网膜病变（DR）筛查的机会有限。人工智能具有提高医疗服务的潜力。因此，我们训练并评估了自动检测 DR 的机器学习算法的诊断准确性：我们纳入了来自 4 个数据库（初级和专业护理机构）的个人彩色眼底照片，排除了无法解读的图像。这些数据集包括巴西患者的图像，这与之前的工作有所不同。这一修改使模型能更有针对性地应用于巴西患者，确保充分捕捉到这一特殊人群的细微差别和特征。样本分为训练样本（70%）和测试样本（30%）。训练卷积神经网络进行图像分类。参考测试是三位眼科医生的综合判定。估算了该算法检测可转诊 DR（中度非增殖性 DR、重度非增殖性 DR、增殖性 DR 和/或有临床意义的黄斑水肿）的灵敏度、特异性和 ROC 曲线下面积：结果：共纳入 15816 张图像（4590 名患者）。任何程度的DR的总发病率为26.5%。与人类评估者（由眼科医生手动诊断 DR 的方法）相比，深度学习算法在检测可转诊 DR 的最高效率点的 ROC 曲线下面积为 0.98（95% CI 0.97-0.98），特异性为 94.6%（95% CI 93.8-95.3），灵敏度为 93.5%（95% CI 92.2-94.9）：一个大型数据库显示，这种深度学习算法能准确检测出可转诊的肺结核。这一发现有助于像巴西这样的全民医疗保健系统优化筛查流程，并可作为改进DR筛查的工具，使其更加灵活，扩大医疗服务的可及性。
期刊介绍：
Diabetology & Metabolic Syndrome publishes articles on all aspects of the pathophysiology of diabetes and metabolic syndrome.
By publishing original material exploring any area of laboratory, animal or clinical research into diabetes and metabolic syndrome, the journal offers a high-visibility forum for new insights and discussions into the issues of importance to the relevant community.
京公网安备 11010802042870号
京ICP备2023020795号-1
分享
求助内容：
应助结果提醒方式：
扫码关注我们
