Evaluation of the effects of an artificial intelligence system on endoscopy quality and preliminary testing of its performance in detecting early gastric cancer: a randomized controlled trial

医学 食管胃十二指肠镜检查 随机对照试验 内窥镜检查 癌症 结肠镜检查 内科学 胃肠病学 外科 结直肠癌
作者
Lianlian Wu,Xinqi He,Mei Liu,Hai Xie,Ping An,Jun Zhang,Heng Zhang,Yaowei Ai,Qiao-Yun Tong,Mingwen Guo,Manling Huang,Cun-Jin Ge,Zhi Yang,Yuan Ji,Jun Li,Wei Zhou,Xiaoda Jiang,Xu Huang,Ganggang Mu,Xinyue Wan,Yanxia Li,Hongguang Wang,Yonggui Wang,Hongfeng Zhang,Di Chen,Dexin Gong,Jing Wang,Li Huang,Jia Li,Liwen Yao,Yijie Zhu,Honggang Yu
出处
期刊:Endoscopy [Georg Thieme Verlag KG]
卷期号:53 (12): 1199-1207 被引量:81
标识
DOI:10.1055/a-1350-5583
摘要

Abstract Background Esophagogastroduodenoscopy (EGD) is a prerequisite for detecting upper gastrointestinal lesions especially early gastric cancer (EGC). An artificial intelligence system has been shown to monitor blind spots during EGD. In this study, we updated the system (ENDOANGEL), verified its effectiveness in improving endoscopy quality, and pretested its performance in detecting EGC in a multicenter randomized controlled trial. Methods ENDOANGEL was developed using deep convolutional neural networks and deep reinforcement learning. Patients undergoing EGD in five hospitals were randomly assigned to the ENDOANGEL-assisted group or to a control group without use of ENDOANGEL. The primary outcome was the number of blind spots. Secondary outcomes included performance of ENDOANGEL in predicting EGC in a clinical setting. Results 1050 patients were randomized, and 498 and 504 patients in the ENDOANGEL and control groups, respectively, were analyzed. Compared with the control group, the ENDOANGEL group had fewer blind spots (mean 5.38 [standard deviation (SD) 4.32] vs. 9.82 [SD 4.98]; P < 0.001) and longer inspection time (5.40 [SD 3.82] vs. 4.38 [SD 3.91] minutes; P < 0.001). In the ENDOANGEL group, 196 gastric lesions with pathological results were identified. ENDOANGEL correctly predicted all three EGCs (one mucosal carcinoma and two high grade neoplasias) and two advanced gastric cancers, with a per-lesion accuracy of 84.7 %, sensitivity of 100 %, and specificity of 84.3 % for detecting gastric cancer. Conclusions In this multicenter study, ENDOANGEL was an effective and robust system to improve the quality of EGD and has the potential to detect EGC in real time.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
大幅提高文件上传限制,最高150M (2024-4-1)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
建议保存本图,每天支付宝扫一扫(相册选取)领红包
实时播报
正直疾完成签到 ,获得积分10
3秒前
10秒前
13秒前
wubiyu完成签到 ,获得积分10
14秒前
16秒前
16秒前
清风完成签到 ,获得积分10
17秒前
19秒前
共享精神应助oui采纳,获得10
22秒前
积极的纸飞机完成签到,获得积分10
22秒前
23秒前
cathy完成签到,获得积分10
42秒前
45秒前
D1fficulty完成签到 ,获得积分10
46秒前
管司机1997完成签到 ,获得积分10
48秒前
50秒前
yolo发布了新的文献求助30
50秒前
在水一方应助学医的小陈采纳,获得10
51秒前
uupp完成签到,获得积分10
54秒前
oracl发布了新的文献求助10
55秒前
彭于晏应助yolo采纳,获得30
58秒前
59秒前
史迪仔发布了新的文献求助10
59秒前
59秒前
an完成签到,获得积分10
59秒前
赘婿应助ycp采纳,获得10
1分钟前
大图图发布了新的文献求助10
1分钟前
高挑的哈密瓜完成签到,获得积分10
1分钟前
chai完成签到,获得积分10
1分钟前
ljq发布了新的文献求助10
1分钟前
学不完了完成签到 ,获得积分10
1分钟前
1分钟前
159完成签到,获得积分10
1分钟前
一叶浮萍归大海完成签到 ,获得积分10
1分钟前
花陵发布了新的文献求助10
1分钟前
lemon完成签到 ,获得积分20
1分钟前
光亮藏鸟发布了新的文献求助10
1分钟前
SOLOMON应助史迪仔采纳,获得10
1分钟前
tuo zhang完成签到,获得积分10
1分钟前
我是老大应助科研通管家采纳,获得10
1分钟前
高分求助中
FILTRATION OF NODULAR IRON WITH CERAMIC FOAM FILTERS 1000
INFLUENCE OF METAL VARIABLES ON THE STRUCTURE AND PROPERTIES OF HEAVY SECTION DUCTILE IRON 1000
Teaching Social and Emotional Learning in Physical Education 900
The Instrument Operations and Calibration System for TerraSAR-X 800
Handbook of Research on Food Processing and Preservation Technologies Volume 3: Computer-Aided Food Processing and Quality Evaluation Techniques 500
A STUDY OF THE EFFECTS OF CHILLS AND PROCESS-VARIABLES ON THE SOLIDIFICATION OF HEAVY-SECTION DUCTILE IRON CASTINGS 500
Filtration of inmold ductile iron 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 有机化学 工程类 生物化学 纳米技术 物理 内科学 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 电极 光电子学 量子力学
热门帖子
关注 科研通微信公众号,转发送积分 2348432
求助须知:如何正确求助?哪些是违规求助? 2054664
关于积分的说明 5115411
捐赠科研通 1785454
什么是DOI,文献DOI怎么找? 891947
版权声明 556871
科研通“疑难数据库(出版商)”最低求助积分说明 475894