Application of artificial intelligence in real-time monitoring of withdrawal speed of colonoscopy

中华消化内镜杂志 Pub Date : 2020-02-20 DOI:10.3760/CMA.J.ISSN.1007-5232.2020.02.010

Xiaoyun Zhu, Lianlian Wu, Suqin Li, Xia Li, Jun Zhang, Shan Hu, Yiyun Chen, Honggang Yu

{"title":"Application of artificial intelligence in real-time monitoring of withdrawal speed of colonoscopy","authors":"Xiaoyun Zhu, Lianlian Wu, Suqin Li, Xia Li, Jun Zhang, Shan Hu, Yiyun Chen, Honggang Yu","doi":"10.3760/CMA.J.ISSN.1007-5232.2020.02.010","DOIUrl":null,"url":null,"abstract":"Objective \nTo construct a real-time monitoring system based on computer vision for monitoring withdrawal speed of colonoscopy and to validate its feasibility and performance. \n \n \nMethods \nA total of 35 938 images and 63 videos of colonoscopy were collected in endoscopic database of Renmin Hospital of Wuhan University from May to October 2018. The images were divided into two datasets, one dataset included in vitro, in vivo and unqualified colonoscopy images, and another dataset included ileocecal and non-cecal area images. And then 3 594 and 2 000 images were selected respectively from the two datasets for testing the deep learning model, and the remaining images were used to train the model. Three colonoscopy videos were selected to evaluate the feasibility of real-time monitoring system, and 60 colonoscopy videos were used to evaluate its performance. \n \n \nResults \nThe accuracy rate of the deep learning model for classification for in vitro, in vivo, and unqualified colonoscopy images was 90.79% (897/988), 99.92% (1 300/1 301), and 99.08% (1 293/1 305), respectively, and the overall accuracy rate was 97.11% (3 490/3 594). The accuracy rate of identifying ileocecal and non-cecal area was 96.70% (967/1 000) and 94.90% (949/1 000), respectively, and the overall accuracy rate was 95.80% (1 916/2 000). In terms of feasibility evaluation, 3 colonoscopy videos data showed a linear relationship between the retraction speed and the image processing interval, which indicated that the real-time monitoring system automatically monitored the retraction speed during the colonoscopy withdrawal process. In terms of performance evaluation, the real-time monitoring system correctly predicted entry time and withdrawal time of all 60 examinations, and the results showed that the withdrawal speed and withdrawal time was significantly negative-related (R=-0.661, P<0.001). The 95% confidence interval of withdrawal speed for the colonoscopy with withdrawal time of less than 5 min, 5-6 min, and more than 6 min was 43.90-49.74, 40.19-45.43, and 34.89-39.11 respectively. Therefore, 39.11 was set as the safe withdrawal speed and 45.43 as the alarm withdrawal speed. \n \n \nConclusion \nThe real-time monitoring system we constructed can be used to monitor real-time withdrawal speed of colonoscopy and improve the quality of endoscopy. \n \n \nKey words: \nQuality control; Artificial intelligence; Colonoscopy; Withdrawal time; Withdrawal speed","PeriodicalId":10072,"journal":{"name":"中华消化内镜杂志","volume":"37 1","pages":"125-130"},"PeriodicalIF":0.0000,"publicationDate":"2020-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"中华消化内镜杂志","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.3760/CMA.J.ISSN.1007-5232.2020.02.010","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

Abstract

Objective To construct a real-time monitoring system based on computer vision for monitoring withdrawal speed of colonoscopy and to validate its feasibility and performance. Methods A total of 35 938 images and 63 videos of colonoscopy were collected in endoscopic database of Renmin Hospital of Wuhan University from May to October 2018. The images were divided into two datasets, one dataset included in vitro, in vivo and unqualified colonoscopy images, and another dataset included ileocecal and non-cecal area images. And then 3 594 and 2 000 images were selected respectively from the two datasets for testing the deep learning model, and the remaining images were used to train the model. Three colonoscopy videos were selected to evaluate the feasibility of real-time monitoring system, and 60 colonoscopy videos were used to evaluate its performance. Results The accuracy rate of the deep learning model for classification for in vitro, in vivo, and unqualified colonoscopy images was 90.79% (897/988), 99.92% (1 300/1 301), and 99.08% (1 293/1 305), respectively, and the overall accuracy rate was 97.11% (3 490/3 594). The accuracy rate of identifying ileocecal and non-cecal area was 96.70% (967/1 000) and 94.90% (949/1 000), respectively, and the overall accuracy rate was 95.80% (1 916/2 000). In terms of feasibility evaluation, 3 colonoscopy videos data showed a linear relationship between the retraction speed and the image processing interval, which indicated that the real-time monitoring system automatically monitored the retraction speed during the colonoscopy withdrawal process. In terms of performance evaluation, the real-time monitoring system correctly predicted entry time and withdrawal time of all 60 examinations, and the results showed that the withdrawal speed and withdrawal time was significantly negative-related (R=-0.661, P<0.001). The 95% confidence interval of withdrawal speed for the colonoscopy with withdrawal time of less than 5 min, 5-6 min, and more than 6 min was 43.90-49.74, 40.19-45.43, and 34.89-39.11 respectively. Therefore, 39.11 was set as the safe withdrawal speed and 45.43 as the alarm withdrawal speed. Conclusion The real-time monitoring system we constructed can be used to monitor real-time withdrawal speed of colonoscopy and improve the quality of endoscopy. Key words: Quality control; Artificial intelligence; Colonoscopy; Withdrawal time; Withdrawal speed

查看原文本刊更多论文

人工智能在结肠镜退出速度实时监测中的应用

目的构建一种基于计算机视觉的结肠镜退出速度实时监测系统，并验证其可行性和性能。方法收集武汉大学人民医院内镜数据库中2018年5月至10月结肠镜检查图像35 938张，视频63段。将图像分为两个数据集，一个数据集包括体外、体内和不合格的结肠镜检查图像，另一个数据集中包括回盲和非盲肠区域图像。然后分别从两个数据集中选择3594和2000幅图像来测试深度学习模型，并用剩余的图像来训练模型。选择3个结肠镜检查视频来评估实时监测系统的可行性，并使用60个结肠镜检测视频来评估其性能。结果深度学习模型对体外、体内和不合格结肠镜图像的分类准确率分别为90.79%（897/988）、99.92%（1300/1 301）和99.08%（1 293/1 305），总体准确率为97.11%（3 490/3 594）。回盲区和非回盲区的识别准确率分别为96.70%（967/1000）和94.90%（949/1000），总准确率为95.80%（1166/2000）。在可行性评估方面，3个结肠镜检查视频数据显示回缩速度与图像处理间隔呈线性关系，这表明实时监测系统在结肠镜退出过程中自动监测回缩速度。在性能评估方面，实时监测系统正确预测了所有60项检查的进入时间和退出时间，结果显示退出速度和退出时间呈显著负相关（R=-0.661，P<0.001）。退出时间小于5分钟、5-6分钟、，6分钟以上分别为43.90-49.74、40.19-45.43和34.89-39.11。因此，39.11被设定为安全退出速度，45.43被设定为警报退出速度。结论所构建的实时监测系统可用于实时监测结肠镜检查的退出速度，提高内镜检查质量。关键词：质量控制；人工智能；结肠镜检查；提款时间；提款速度

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

中华消化内镜杂志

CiteScore

0.10

自引率

0.00%

发文量

7555

期刊介绍： Chinese Journal of Digestive Endoscopy is a high-level medical academic journal specializing in digestive endoscopy, which was renamed Chinese Journal of Digestive Endoscopy in August 1996 from Endoscopy. Chinese Journal of Digestive Endoscopy mainly reports the leading scientific research results of esophagoscopy, gastroscopy, duodenoscopy, choledochoscopy, laparoscopy, colorectoscopy, small enteroscopy, sigmoidoscopy, etc. and the progress of their equipments and technologies at home and abroad, as well as the clinical diagnosis and treatment experience. The main columns are: treatises, abstracts of treatises, clinical reports, technical exchanges, special case reports and endoscopic complications. The target readers are digestive system diseases and digestive endoscopy workers who are engaged in medical treatment, teaching and scientific research. Chinese Journal of Digestive Endoscopy has been indexed by ISTIC, PKU, CSAD, WPRIM.