Impact of a real-time automatic quality control system on colorectal polyp and adenoma detection: a prospective randomized controlled study (with videos)

Su, Jianping; Li, Zhen; Shao, Xuejun; Ji, Chaoran; Ji, Rui; Zhou, Ran; Li, Guangchao; Liu, Guan-Qun; He, Yi-Shan; Zuo, Xu; Li, Yanqing

doi:10.1016/j.gie.2019.08.026

Cited by 219 publications

(269 citation statements)

References 30 publications

Supporting

Mentioning

262

Contrasting

Order By: Relevance

“…The blind spot rate in a randomized controlled trial with 324 patients was significantly reduced in the WISENSE group (5.86%) compared with the controls (22.46%). In the field of colonoscopy, Su et al developed an automatic quality control system (AQCS) for timing withdrawal of the colonoscopy, supervising stability of withdrawal inspection, evaluating bowel preparation, and detecting colorectal polyps, as assessed in a prospective randomized controlled trial. The results indicate that AQCS can increase the adenoma detection rate (0.289 vs 0.165).…”

Section: Current Situation Of Ai‐aided Endoscopic Image Recognitionmentioning

confidence: 99%

“…It is possible that AI could be used to mitigate variations in endoscopists’ technical skills, including the thoroughness of the endoscopist's investigations and examination time, and thus improving the quality of the daily endoscopy. The CADm has thus great potential for training junior physicians and improving their performance in real clinical practice . In the future, smart CNN‐based systems combining lesion detection, characterization, and quality control should be designed for globally improving the performance of endoscopists.…”

Section: Current Situation Of Ai‐aided Endoscopic Image Recognitionmentioning

confidence: 99%

“…For repetitive and time‐consuming tasks, such as capsule endoscopy, AI far outperforms doctors . Moreover, CADm can carry out on the real‐time quality control of the endoscopy, reminding physicians of the standards required in such operation . As assistant observers or clinical decision‐makers, AI‐aided systems undoubtedly have a promising future in increasing the effective training of junior endoscopists .…”

Section: Future Challenges Of Ai In Gi Endoscopymentioning

confidence: 99%

“…Moreover, CADm can carry out on the real‐time quality control of the endoscopy, reminding physicians of the standards required in such operation . As assistant observers or clinical decision‐makers, AI‐aided systems undoubtedly have a promising future in increasing the effective training of junior endoscopists . Furthermore, it has emerged that, given the advantages of computer vision, the depth of tumor invasion can be predicted …”

Section: Future Challenges Of Ai In Gi Endoscopymentioning

confidence: 99%

“…In this situation, the accuracy of CAD might be exaggerated, as its good performance when using static, high‐quality images does not guarantee that recognition will be successful when using dynamic videos. Thus, unin‐formative frames should be included in the models and differentiated through auxiliary modules, making the system more robust . To facilitate real‐time assessments during endoscopic examinations, training or testing CNN with unprocessed videos should be used to make the model more realistic .…”

Section: Future Challenges Of Ai In Gi Endoscopymentioning

confidence: 99%

See 4 more Smart Citations

Application of artificial intelligence in gastrointestinal endoscopy

et al. 2019

J of Digest Diseases

Self Cite

View full text Add to dashboard Cite

With recent significant improvements in artificial intelligence (AI), especially in the field of deep learning, an increasing number of studies have evaluated the use of AI in endoscopy to detect and diagnose gastrointestinal (GI) lesions. The present review summarizes current publications on the use of AI in GI endoscopy and focuses on the challenges and future of AI-aided systems. We expect AI to provide an effective and practical method for endoscopists in lesion detection and characterization as well as in quality control in endoscopy. However, so far, most studies have remained at the preclinical stage. More attention should be paid in the future to the use of AI in reallife clinical applications.Artificial intelligence (AI), first proposed in the 1950s, 1 uses computers to simulate certain thought processes and forms of intelligent behavior. In recent years, one way to achieve AI, namely machine learning, has been used in a variety of medical fields, such as radiology, pathology, ophthalmology, orthopedics, obstetrics, and gynecology, in an attempt to improve physicians' diagnoses. 2 In the field of gastrointestinal (GI) endoscopy endoscopists face heavy workloads, and different kinds of endoscopic images, such as those of esophagogastroduodenoscopy, colonoscopy and capsule endoscopy, has led specialty gastroenterology societies to call for practical and effective methods in order to improve the efficiency and quality of endoscopy in daily clinical practice.With the advent of data-driven deep learning (DL), the inefficient and incomplete feature extraction that results from manual extraction by using traditional machine learning has been thoroughly addressed, and this has revolutionized the research and development of AI. 3,4 After having been trained by a large number of labeled images, the DL algorithm can classify enormous quantities of images. 3 A variety of DL algorithms have thus been created, of which convolutional neural networks (CNN) are the most widely used for medical image recognition.However, DL has some defects, including the requirement of highquality databases and its black-box nature, which have been summarized in Table 1. AI technology used in GI endoscopy falls mainly in two groups:computer-aided detection (CADe) system and computer-aided diagnosis (CADx) system. 5 CADe system is designed to detect and track GI lesions, while CADx system focuses mainly on identifying and characterizing them. Moreover, computer-aided monitoring (CADm) systems are designed for evaluating the examination procedure and † These authors contributed equally to this article.

show abstract

Section: Current Situation Of Ai‐aided Endoscopic Image Recognitionmentioning

confidence: 99%

Section: Current Situation Of Ai‐aided Endoscopic Image Recognitionmentioning

confidence: 99%

Section: Future Challenges Of Ai In Gi Endoscopymentioning

confidence: 99%

Section: Future Challenges Of Ai In Gi Endoscopymentioning

confidence: 99%

Section: Future Challenges Of Ai In Gi Endoscopymentioning

confidence: 99%

See 3 more Smart Citations

Application of artificial intelligence in gastrointestinal endoscopy

et al. 2019

J of Digest Diseases

Self Cite

View full text Add to dashboard Cite

show abstract

Randomized Clinical Trials of Machine Learning Interventions in Health Care

et al. 2022

View full text Add to dashboard Cite

ImportanceDespite the potential of machine learning to improve multiple aspects of patient care, barriers to clinical adoption remain. Randomized clinical trials (RCTs) are often a prerequisite to large-scale clinical adoption of an intervention, and important questions remain regarding how machine learning interventions are being incorporated into clinical trials in health care.ObjectiveTo systematically examine the design, reporting standards, risk of bias, and inclusivity of RCTs for medical machine learning interventions.Evidence ReviewIn this systematic review, the Cochrane Library, Google Scholar, Ovid Embase, Ovid MEDLINE, PubMed, Scopus, and Web of Science Core Collection online databases were searched and citation chasing was done to find relevant articles published from the inception of each database to October 15, 2021. Search terms for machine learning, clinical decision-making, and RCTs were used. Exclusion criteria included implementation of a non-RCT design, absence of original data, and evaluation of nonclinical interventions. Data were extracted from published articles. Trial characteristics, including primary intervention, demographics, adherence to the CONSORT-AI reporting guideline, and Cochrane risk of bias were analyzed.FindingsLiterature search yielded 19 737 articles, of which 41 RCTs involved a median of 294 participants (range, 17-2488 participants). A total of 16 RCTS (39%) were published in 2021, 21 (51%) were conducted at single sites, and 15 (37%) involved endoscopy. No trials adhered to all CONSORT-AI standards. Common reasons for nonadherence were not assessing poor-quality or unavailable input data (38 trials [93%]), not analyzing performance errors (38 [93%]), and not including a statement regarding code or algorithm availability (37 [90%]). Overall risk of bias was high in 7 trials (17%). Of 11 trials (27%) that reported race and ethnicity data, the median proportion of participants from underrepresented minority groups was 21% (range, 0%-51%).Conclusions and RelevanceThis systematic review found that despite the large number of medical machine learning–based algorithms in development, few RCTs for these technologies have been conducted. Among published RCTs, there was high variability in adherence to reporting standards and risk of bias and a lack of participants from underrepresented minority groups. These findings merit attention and should be considered in future RCT design and reporting.

show abstract