相关实验视频

Updated: Jan 14, 2026

Using Retinal Imaging to Study Dementia

Using Retinal Imaging to Study Dementia

Published on: November 6, 2017

通过深度学习通过视网膜底部图像预测晚期慢性病

Chuan-Fan Hsu¹, Tung-Min Yu^2,3,4, Ya-Lun Wu¹

¹Artificial Intelligence and Robotics Innovation Center, China Medical University Hospital, Taichung, Taiwan.

Scientific reports

|October 24, 2025

概括

使用视网膜图像的深度学习模型可以检测慢性病 (CKD). 最好的模型结合了双边视网膜图像,并实现了0.868AUC,优于单图像方法.

科学领域:

眼科医生眼科眼科
腎臟病學 (nephrology) 是一種醫學專業.
人工智能的人工智能

背景情况:

慢性病 (CKD) 是一个重要的全球健康问题.
早期发现CKD对于有效管理和改善患者的治疗结果至关重要.
视网膜底部图像为系统健康提供了一个非侵入性的窗口,可能会揭示CKD的迹象.

研究的目的:

开发和评估一种深度学习模型,用于使用视网膜底图像检测CKD.
为了比较不同的模型配置,包括单对双边图像和人口统计数据集成.
评估各种深度学习架构和用于CKD检测的培训策略的性能.

主要方法:

利用了来自17442名患者的42,963张视网膜底图像的大型数据集,估计测量了淋巴细胞过率 (eGFR).
开发了三种深度学习模型配置:单图像 (模型A),单图像与人口统计 (模型B) 和双边图像 (模型C).
对比了EfficientNet-B3和EfficientNetV2-S架构,以及单模型与5倍交叉验证 (CV) 组合培训策略.

主要成果:

使用EfficientNet-B3和5倍CV组合的双边图像模型 (模型C) 获得了最高的性能 (AUC 0.868,灵敏度 0.792,特异性 0.788).
这种整体策略在统计学上优于单一模型方法 (AUC 0.850,p < 0.001).
在单个模型中,B模型显示了最高的AUC (0.857) 和灵敏度 (0.794),而C模型则提供了最高的特异性 (0.799).

结论:

使用视网膜底图像的深度学习模型显示出检测慢性病的巨大潜力.
整合双边视网膜图像和使用整体策略可以提高用于CKD检测的模型性能.
开发的模型在识别晚期CKD方面表现出卓越的表现,特别是在糖尿病患者中.

关键词:

慢性脏疾病慢性脏疾病深度学习是一种深度学习.视网膜底部图像视网膜底部图像

更多相关视频

Author Spotlight: Unraveling the Pathogenesis of Age-Related Macular Degeneration and Discovering Potential Therapies

Author Spotlight: Unraveling the Pathogenesis of Age-Related Macular Degeneration and Discovering Potential Therapies

Published on: July 28, 2023

5/6 Nephrectomy Using Sharp Bipolectomy Via Midline Laparotomy in Rats

5/6 Nephrectomy Using Sharp Bipolectomy Via Midline Laparotomy in Rats

Published on: April 4, 2025

相关概念视频

Chronic Kidney Disease I: Introduction

Chronic Kidney Disease I: Introduction

Chronic Kidney Disease (CKD) arises when the kidneys progressively lose their ability to function, ultimately leading to end-stage renal disease. At this advanced stage, the kidneys can no longer filter waste or maintain essential body functions, requiring renal replacement therapy (RRT) through dialysis or a kidney transplant for survival.Early-stage chronic kidney disease and detection challengesIn CKD's early stages, symptoms often remain absent because healthy nephrons compensate for...

Chronic Kidney Disease III: Interprofessional Care

Chronic Kidney Disease III: Interprofessional Care

Chronic kidney disease (CKD) requires collaborative and comprehensive management. CKD progresses through stages and can lead to end-stage kidney disease (ESKD) if untreated. Interprofessional collaboration and patient education are crucial, enabling patients to manage their health and improve their quality of life.Diagnostic approach for chronic kidney diseaseThe diagnosis of CKD primarily focuses on the glomerular filtration rate (GFR), which assesses kidney function by measuring how well...

Drug Dosing in Renal Diseases: Estimation of Glomerular Filtration Rate Based on Serum Creatinine Concentration

Drug Dosing in Renal Diseases: Estimation of Glomerular Filtration Rate Based on Serum Creatinine Concentration

Glomerular filtration rate (GFR) can be estimated from serum creatinine using the modification of diet in renal disease (MDRD) formula or the chronic kidney disease–epidemiology collaboration (CKD–EPI) equation. Both methods are widely used in clinical practice to assess kidney function and guide treatment decisions.The MDRD equation does not require weight or height measurements and is normalized to the body surface area of 1.73 m², considered the average adult surface area.