Skip Navigation
Skip to contents

Clin Endosc : Clinical Endoscopy

Search
Advanced Search
OPEN ACCESS
mobile search button mobile menu button

Articles

Page Path
HOME > Clin Endosc > Ahead-of print articles > Article
Review Computer-aided quality control in colonoscopy: clinical applications and limitations
Elizabeth Lee Yoong Chen1,2orcid, James Weiquan Li1,2orcid

DOI: https://doi.org/10.5946/ce.2025.309
Published online: December 17, 2025

1Gastroenterology and Hepatology Service, Department of General Medicine, Sengkang General Hospital, Singapore, Singapore

2Academic Medical Centre, Duke-NUS, Singapore, Singapore

Correspondence: James Weiquan Li Gastroenterology and Hepatology Service, Department of General Medicine, Sengkang General Hospital, 110 Sengkang East Way, Singapore 544886, Singapore E-mail: james.li.w.q@singhealth.com.sg
• Received: August 30, 2025   • Revised: October 18, 2025   • Accepted: October 23, 2025

© 2025 Korean Society of Gastrointestinal Endoscopy

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

  • 1,191 Views
  • 226 Download
  • 1 Web of Science
  • 1 Crossref
Full Article

Download PDF Download PDF

  • Computer-aided quality control (CAQ) systems are redefining colonoscopy by enabling the objective evaluation of procedural metrics and providing real-time feedback. This review explores the clinical utility, implementation barriers, and future prospects of CAQ, with an emphasis on its role in standardizing quality assessment and enhancing patient outcomes. A systematic search of PubMed (inception to January 2025) identified 66 relevant publications, including eight systematic reviews or meta-analyses, seven randomized controlled trials, and five cohort studies, in addition to validation and observational reports. CAQ systems improve traditional quality indicators such as withdrawal time, bowel preparation scores, and cecal intubation rates (CIRs). Emerging metrics—including effective withdrawal time, fold examination quality, and withdrawal speed—offer novel, quantifiable insights. Artificial intelligence-assisted colonoscopy consistently increases adenoma detection rates (from 38.5% to 47.9%) and extends withdrawal time (from 5.68 to 7.03 minutes). Automated systems achieve high accuracy in bowel preparation scoring (93.3%), cecal intubation recognition (95.5%), and surveillance interval assignment (92.0%), thereby addressing persistent gaps in documentation and follow-up care. CAQ systems hold transformative promise for improving colonoscopy quality. Addressing implementation challenges—including false positives, clinician adoption, cost, and regulatory issues—is essential. Future research should emphasize comparative effectiveness, standardized metrics, and large-scale clinical integration to help reduce the burden of colorectal cancer.
  • Keywords: Artificial intelligence; Colonoscopy; Quality control
The application of artificial intelligence (AI) to quality control in colonoscopy has expanded considerably over the past decade, with advances in deep learning creating numerous opportunities—from improving detection1,2 and diagnostic accuracy3-6 to optimizing the efficiency of clinical workflows and refining endoscopic skills.7,8 Collectively, these developments hold promise for reducing the incidence of colorectal cancer, which is the third most common cancer worldwide and the second most common cause of cancer-related mortality.9 Efforts to improve the early detection of neoplastic colonic lesions have been, and remain, essential for enabling early diagnosis and timely treatment. Consequently, it is unsurprising that significant research has focused on computer-aided quality control (CAQ) in endoscopy to better understand how AI can be leveraged and how its clinical applications can be optimized in this field, in parallel with technological advancements in other industries beyond medicine.
This narrative review aims to explore the clinical applications of CAQ in colonoscopy, extending beyond its utility in detection and diagnosis to its broader integration into clinical practice. It also highlights both conventional and emerging quality control domains and identifies current gaps and limitations within this evolving area of research.
A literature search was conducted in the PubMed database from inception to January 21, 2025, using relevant search terms combined with the Boolean operators “AND” and “OR”. The search terms included “computer-aided,” “artificial intelligence”, “deep learning”, “colonoscopy”, “quality control”, “performance measures”, “adenoma detection rate”, “clinical applications”, and “performance evaluation”. Additional relevant articles were identified through manual searches of reference lists from related publications.
This process yielded a total of 66 studies, comprising 8 systematic reviews or meta-analyses, 7 randomized controlled trials (RCTs), 5 cohort studies, one consensus position paper, and 12 additional reports—including validation and observational studies—that provided insights for this review.
Indicators of quality in colonoscopy: conventional and new
Conventional quality indicators in colonoscopy, as outlined in the American Gastroenterological Association Clinical Practice Update in 2021, include adequate bowel preparation—defined as a Boston bowel preparation scale (BBPS) score ≥6, with each segment score ≥2—achieved in ≥90% of screening and surveillance colonoscopies; a CIR of at least 90%; a withdrawal time of at least 6 minutes; an adenoma detection rate (ADR) of at least 30%; and a serrated lesion detection rate of at least 7%.10 In addition, appropriate post-polypectomy surveillance was emphasized as one of the seven key performance measures identified by the 2017 European Society of Gastrointestinal Endoscopy Quality Improvement Initiative.11
The traditional method of assessing and measuring quality metrics—typically conducted through institutional audits of ADR and individual assessments by endoscopists, followed by documentation of indicators such as BBPS and withdrawal time—is subject to variability in observation and documentation among practitioners and across institutions, and it may not accurately reflect the true quality of the procedure. Moreover, these data are often reviewed retrospectively and therefore cannot provide real-time feedback to the operator during the colonoscopy.
With the advent of AI, new parameters in addition to conventional quality measures—such as withdrawal speed and fold examination quality (FEQ)—have been introduced. Both conventional and novel quality metrics can now be measured more objectively through CAQ systems, which also enable real-time feedback to enhance procedural quality at the point of care. Another area where AI may play an increasingly important role is in promoting the standardization of documentation and reporting, which ultimately supports the assignment of appropriate surveillance intervals.
AI tools for quality control in endoscopy
CAQ systems in colonoscopy utilize AI-based tools to monitor and assess quality metrics in real time, including CIR, withdrawal time, bowel preparation score, and FEQ. Notable examples, such as the ENDOANGEL and WISENSE systems (Table 1),12-15 are discussed in this narrative review. These systems represent a paradigm shift toward objective and standardized quality assessment. Figure 1 presents a schematic illustration of how CAQ systems can be integrated into real-time colonoscopy within clinical practice.
Furthermore, integrating AI with electronic medical records facilitates standardized documentation through automated report generation, linking procedural findings with patient records, and assigning appropriate surveillance intervals post-colonoscopy. This integration addresses one of the most significant challenges in colonoscopy quality control—maintaining consistent and accurate documentation across diverse healthcare settings and providers.
Although some of these tools are currently more accessible for clinical use than others, most hold considerable potential to shape how AI influences quality control in future surveillance and screening colonoscopies. This review focuses on CAQ systems in colonoscopy and their potential clinical applications.
Role of AI in optimizing conventional quality indicators
CAQ systems have the potential to enhance objectivity in assessment and documentation while improving conventional quality indicators such as withdrawal time, BBPS scores, and CIR. AI-assisted colonoscopy has been shown to increase withdrawal time, which appears to correlate with improved ADRs. In a prospective RCT, Su et al.12 demonstrated that an automatic quality control system (AQCS)—developed to measure withdrawal time, monitor scope stability during withdrawal, evaluate bowel preparation, and detect colonic polyps—significantly increased both ADR and the mean number of adenomas detected per procedure. The use of AQCS during colonoscopy was also associated with longer withdrawal times compared to the control group (7.03 minutes vs. 5.68 minutes).
These results were consistent with a single-center retrospective study conducted in New Zealand, which compared artificial intelligence-assisted colonoscopy (AIAC) with conventional colonoscopy and reported a significantly longer withdrawal time (15 minutes vs. 13 minutes) and a higher ADR (47.9% vs. 38.5%) in the AIAC group compared to conventional colonoscopy group.13
Furthermore, the advantage of using AI systems to measure withdrawal time lies in their ability to calculate the effective withdrawal time—defined as the proportion of effective examination frames (i.e., frames providing a clear view of the colonic wall)—which accounts for both withdrawal speed and mucosal visualization quality.14 This novel metric enables accurate and objective evaluation of the time effectively spent examining the colon during withdrawal, which has been shown to correlate with increased ADRs.15
CAQ systems have also been developed to provide objective assessments of bowel preparation quality in a reliable and reproducible manner, thereby facilitating feedback and improving the quality of endoscopic examinations. One such system, ENDOANGEL—a novel AI-based tool for evaluating bowel preparation—demonstrated 93.3% accuracy in assessing bowel preparation scores.16 Additionally, the AI-based open-source automatic bowel preparation scale established a cutoff value of >0.09 for identifying colonoscopies that should be considered for re-examination, enabling automatic and objective assessment of bowel preparation quality in real time.17 AI-assisted CAQ systems also allow adjustments to be made during colonoscopy to meet quality standards. Moreover, by providing endoscopists with continuous opportunities to reflect on and refine their examination technique with each AI-assisted procedure, these systems naturally contribute to the progressive enhancement of endoscopic skills over time.
AI-assisted documentation of anatomical landmarks, particularly, enables providers to track the completeness of colonoscopy and represents an important quality indicator that has been shown to correlate with higher ADRs.18 Deep learning algorithms have been developed to automate landmark documentation during colonoscopy. For instance, in a development and validation study, Chang et al.19 reported 95.5% accuracy in determining CIR and 94.6% accuracy in differentiating bowel preparation status when the model was applied to real-world colonoscopy images and reports.
These aspects of CAQ hold promise for implementation across diverse clinical settings, paving the way toward standardized quality assurance in colonoscopy and contributing to improved quality benchmarks on a broader scale.
Role of AI in measuring new parameters in quality assessment
To further explore how CAQ systems may enhance quality indicators, it is appropriate to highlight novel parameters that, before the advent of AI-based systems, could not be quantified—namely, withdrawal speed, scope stability, and FEQ.
Although withdrawal time has been recognized as a key quality indicator, a closely related yet more difficult-to-measure parameter is withdrawal speed. Maintaining a stable and consistent withdrawal speed allows for a more thorough examination of the colonic mucosa and places the endoscopist in a favorable position to detect pathology.14
In 2012, Filip et al.20 conducted a pilot feasibility and technical validation study that tested a novel real-time system, Colometer, which provided feedback on image clarity, bowel preparation, and withdrawal speed, thereby laying the foundation for subsequent validation and development of similar systems. Since then, several AI-assisted quality systems have incorporated real-time withdrawal speed monitoring into their algorithms, including Chen et al.’s EfficientNetB2 model21 and ENDOANGEL. The latter, in addition to its ability to monitor withdrawal speed, offers the advantage of alerting endoscopists to blind spots caused by endoscope slippage.22 This capability is achieved through calculation of the proportion of overspeed frames (POF) during colonoscopy withdrawal—defined as the percentage of frames with a withdrawal speed exceeding 44 mm/s—after filtering out frames associated with biopsies or therapeutic procedures.14 Gong et al.22 found that POF was inversely associated with ADR, suggesting its potential as a novel quality metric. The overall effectiveness of ENDOANGEL in improving adenoma yield was demonstrated by a significantly higher ADR in the ENDOANGEL group compared to controls (16% vs. 8%).
With regard to scope stability during withdrawal, Su et al.12’s AQCS was able to monitor and provide feedback on this metric in real time, prompting endoscopists to adjust and regulate their inspection speed and duration while examining each colonic flexure and fold. This resulted in significantly longer withdrawal times and higher ADRs in the AQCS group.
Improved scope stability and slower withdrawal speeds are associated with enhanced FEQ—a metric that has become measurable only with CAQ systems. Thakkar et al.23’s novel AI-based system provided real-time feedback to endoscopists on four key quality-of-examination parameters: surface area, distension, bowel preparation, and mucosal visualization clarity. Each parameter was scored for every colonic segment, enabling endoscopists to review and refine their performance and the adequacy of their examinations in real time.
In addition, Liu et al.15’s prospective observational study developed and validated an AI-based system for assessing FEQ. The system’s evaluation of FEQ showed strong correlations with ADR, withdrawal time, and expert-assigned FEQ scores, and AI assistance improved FEQ among endoscopists with previously low ADRs. Although this study was limited by a small sample size and did not establish a clinically relevant FEQ threshold, it laid the groundwork for larger RCTs to confirm the system’s impact on FEQ and ADR before clinical implementation.
AI-assisted documentation in ensuring appropriate surveillance intervals
Peterson et al.24 developed a machine learning–based system that automatically extracted colonoscopy findings and classified patients into appropriate risk categories with corresponding surveillance intervals, achieving an overall accuracy of 92% in assigning appropriate follow-up intervals. In addition, an automated surveillance system developed by Wu et al.25 improved physician adherence to post-colonoscopy surveillance guidelines and reduced workload among physicians and nurses. Together, these studies underscore the role of AI-assisted documentation in ensuring appropriate surveillance intervals and, consequently, enhancing the quality of colorectal cancer surveillance programs.
The recent Asia-Pacific Consensus on the Use of Artificial Intelligence in Colorectal Cancer Screening and Surveillance highlighted the importance of CAQ in colonoscopy reporting and its role in standardizing reporting quality.26 By minimizing variations in colonoscopy report quality across different healthcare settings, AI-assisted documentation has the potential to improve the accuracy and completeness of reporting, thereby enabling the development of refined thresholds for quality assessment. Such standardization may ultimately translate into more consistent and appropriate post-colonoscopy surveillance recommendations.
Role of AI in improving the training and skill set of endoscopists
Another potential clinical application of CAQ systems lies in the training and refinement of endoscopic skills. These systems enhance learning primarily through real-time feedback on FEQ, withdrawal time and speed, and scope stability, in addition to computer-aided detection (CADe) systems such as GI Genius (Medtronic).
Interestingly, a prospective observational study by Okumura et al.27 investigating the long-term effects of CADe implementation on endoscopists demonstrated a sustained improvement in adenoma detection among those with a high baseline ADR.27 Notably, this study found no evidence of deskilling, addressing concerns that the integration of AI could lead to overreliance and skill deterioration among endoscopists over time.
The influence of AI on endoscopy training has also been demonstrated by studies involving trainees. A single-center retrospective cohort study by Khouri et al.28 reported a statistically significant increase in serrated polyp detection rates (SDR) with the use of GI Genius among trainee fellows. Similarly, a prospective observational study by Orzeszko et al.29 showed that AI-trained endoscopists exceeded aspirational quality benchmarks, with longer withdrawal times and higher SDRs compared to those trained through conventional methods. Furthermore, a prospective study by Yamaguchi et al.30 evaluating the CAD EYE (Fujifilm) system among gastroenterology trainees found reduced adenoma miss rates and improved accuracy in adenoma localization and identification.
These findings highlight the potential of AI-assisted colonoscopy as a transformative tool in endoscopy training. By providing objective, real-time feedback, standardizing assessment, enabling repetitive practice with personalized guidance, and facilitating longitudinal monitoring of skill development, AI has the capacity to bridge the gap between theoretical knowledge and expert-level performance. This integration not only accelerates learning among trainees but also contributes to long-term improvements in colonoscopy quality standards.
Healthcare system integration
The successful implementation of CAQ systems requires thoughtful consideration of existing healthcare infrastructure and clinical workflows. Different healthcare settings present unique challenges in terms of technological readiness, staff training needs, and compatibility with existing endoscopy systems. Large tertiary centers may have the resources and technical expertise to adopt advanced AI technologies, whereas smaller community hospitals may encounter significant barriers such as limited information technology support, budgetary constraints, and resistance to workflow modifications.
The integration process typically unfolds in several stages: system installation and calibration, staff training programs, pilot testing, and gradual deployment across departments. Each stage presents distinct challenges, including ensuring hardware and software compatibility with existing endoscopy equipment and designing comprehensive training modules that accommodate varying levels of technological proficiency among healthcare personnel.
Training requirements for healthcare professionals
The effective adoption of CAQ systems requires comprehensive training programs for endoscopists, nursing staff, and technical personnel. These training programs must extend beyond basic system operation to include the interpretation of AI-generated alerts, understanding of quality metrics, and maintenance of sound clinical judgment when AI recommendations differ from clinical assessment.
Endoscopists must acquire proficiency in interpreting real-time feedback from CAQ systems while maintaining procedural focus. This balance demands skill in leveraging AI support without becoming overly dependent on automated guidance. Accordingly, training initiatives should emphasize the complementary role of AI tools in augmenting, rather than replacing, clinical expertise.
Nursing staff and endoscopy technicians also require instruction in system maintenance, troubleshooting common technical issues, and understanding quality indicators to provide optimal support to endoscopists. The creation of standardized training curricula and certification programs is essential to ensure uniform implementation and competency across diverse healthcare environments.
Regulatory considerations and standards
The regulatory landscape for AI-based medical devices continues to evolve, with different jurisdictions adopting diverse approaches to approval and oversight. In the United States, the Food and Drug Administration has established pathways for AI-enabled medical technologies, including frameworks for software as a medical device. The European Union’s Medical Device Regulation offers similar oversight mechanisms, whereas other regions are in the process of developing their own regulatory frameworks.
CAQ systems must demonstrate both safety and efficacy through rigorous clinical evaluation before obtaining regulatory approval. This process typically involves extensive validation studies, comprehensive risk assessments, and adherence to post-market surveillance requirements. The adaptive nature of AI algorithms—particularly those that continuously learn and update over time—poses unique regulatory challenges in maintaining long-term safety, performance consistency, and clinical effectiveness.
Quality standards for CAQ systems should encompass accuracy benchmarks, acceptable false-positive rates, system reliability, and robust cybersecurity safeguards. International standardization bodies are actively developing harmonized guidelines to promote global adoption while ensuring patient safety, data protection, and interoperability across healthcare systems.
Advanced AI architectures
The future of CAQ systems lies in the development of more advanced AI architectures capable of providing comprehensive quality assessments across multiple parameters simultaneously. Emerging technologies such as transformer networks and attention mechanisms show significant potential to enhance the accuracy and reliability of quality evaluations while reducing false-positive rates.
Multi-modal AI systems that integrate visual data with complementary information sources—such as patient medical history, previous colonoscopy results, and real-time physiological parameters—represent a major frontier in CAQ development. These systems could enable the generation of personalized quality benchmarks and recommendations tailored to individual patient risk factors and clinical characteristics.
Predictive analytics and remote quality monitoring
Future CAQ systems may incorporate predictive analytics to identify patients at increased risk of missed lesions or suboptimal examinations before the procedure begins. By analyzing patient characteristics, prior colonoscopy findings, and real-time quality indicators, these systems could generate personalized recommendations for examination protocols and follow-up intervals.
The integration of CAQ systems with telemedicine platforms could facilitate remote quality monitoring and real-time expert consultation during colonoscopy procedures. This capability would be particularly valuable in resource-limited environments where access to experienced endoscopists is constrained. Furthermore, remote quality assessment could support continuous professional development and quality improvement initiatives across diverse healthcare settings.
Limitations of CAQ in colonoscopy
Despite the promising outlook of computer-aided systems in shaping the landscape of quality control in colonoscopy, several limitations hinder their widespread adoption in clinical practice.
Although CADe systems have been integrated into clinical settings, and numerous studies have demonstrated their effectiveness in improving lesion detection rates,31-34 evidence suggests that much of this improvement stems from increased detection of diminutive polyps (<5 mm).35-37 This observation raises questions regarding the clinical significance of AI-enhanced detection and its cost-effectiveness. Specifically, does AI meaningfully contribute to identifying advanced adenomas, or does it primarily increase the detection of diminutive lesions, thereby potentially adding to the burden on healthcare resources?38
Another key limitation of CADe systems is the occurrence of false positives, which may contribute to operator distraction and fatigue. A study by Zhang et al.39 demonstrated that a higher ADR correlated with a lower rate of false positives per minute (FPPM) and that the beneficial effect of CADe diminished as FPPM increased.
In a retrospective analysis by Nehme et al.,40 a survey evaluating attitudes toward AI-assisted colonoscopy revealed mixed perspectives. The most common concerns included elevated false-positive signals (82.4%), increased procedural distraction (58.8%), and the perception of prolonged procedure time (47.1%). Moreover, the study found no improvement in ADR among endoscopists with an already high baseline ADR during routine use of AI-assisted colonoscopy in clinical practice.
These findings contribute to the heterogeneous attitudes and reduced readiness among endoscopists40 to adopt AI-assisted approaches in colonoscopy, representing a key barrier to the uptake of CAQ systems. For large-scale implementation to succeed, further multicenter studies are required across varied clinical settings to demonstrate consistent improvements in quality outcomes. Only with such evidence can attitudes shift, thereby enhancing receptiveness toward AI-assisted quality improvement in colonoscopy across healthcare environments.
This aligns with the 2025 Asia-Pacific Consensus statement, which emphasized the importance of engaging both endoscopists and endoscopy assistants to build trust and facilitate the successful integration of AI-assisted colonoscopy into colorectal cancer screening practice.26
In addition to existing gaps in the evidence surrounding CAQ systems in colonoscopy, another major barrier to their widespread implementation is the lack of standardized protocols and the absence of head-to-head or comparative effectiveness trials evaluating different CAQ systems. Such studies are essential to determine which systems are most suitable for various clinical contexts—such as differing patient populations, geographical regions, and endoscopist experience levels—thereby informing evidence-based adoption and investment decisions.
Furthermore, most validation and evaluation studies of CAQ systems have excluded patients with underlying chronic bowel diseases or altered anatomy, such as those with inflammatory bowel disease or a history of bowel resection. This exclusion limits the generalizability and clinical applicability of CAQ systems to specialized patient populations.
Economic considerations and cost-effectiveness
Economic factors play a critical role in determining the feasibility of large-scale implementation of AI technologies in routine colonoscopy screening programs. Cost-effectiveness analyses are therefore essential for evaluating the practicality of integrating such systems across different healthcare environments and geographical regions.
A United States–based modeling study by Areia et al.41 found that the use of AI-assisted detection tools could prevent 7,194 additional colorectal cancer cases and 2,089 cancer-related deaths annually, while generating estimated savings of 290 million United States dollars per year. The authors concluded that AI integration in screening colonoscopy represents a cost-saving strategy for reducing colorectal cancer incidence and mortality. However, cost-effectiveness remains context-dependent. While the financial investment may be justified in high-resource settings where improved colonoscopy quality translates to long-term cancer prevention, the same may not hold true in low-resource environments, where the incremental benefits of AI may be insufficient to offset the high initial and maintenance costs.
Economic evaluation of CAQ systems must account not only for the direct expenses of acquisition and maintenance but also for indirect costs associated with staff training, workflow adaptation, and potential increases in procedural duration. Conversely, the economic benefits extend beyond immediate quality enhancement to include reduced healthcare expenditure through cancer prevention, fewer repeat procedures due to improved examination adequacy, and greater operational efficiency from standardized quality assessment.
Potential legal considerations for CAQ systems in colonoscopy
An important consideration in integrating computer-aided systems into colonoscopy practice is the issue of medicolegal liability. Currently, clinicians remain legally accountable for diagnostic and procedural outcomes, irrespective of whether AI tools are used in the process. As AI becomes increasingly embedded within clinical and procedural workflows, questions arise as to whether legal responsibility should, in certain cases, extend to AI developers, vendors, or manufacturers—particularly when diagnostic errors lead to suboptimal patient outcomes. In the context of CAQ systems in colonoscopy, AI functions primarily as a real-time quality-control adjunct; therefore, the physician is expected to retain ultimate decision-making authority. Nevertheless, evolving medicolegal frameworks will be necessary to clarify accountability and ensure legal certainty as AI integration in colonoscopy continues to advance.
Research priorities and evidence gaps
Several critical research priorities emerge from the current literature on CAQ systems in colonoscopy. Large-scale, multicenter RCTs are required to establish the clinical effectiveness of various CAQ systems across diverse patient populations and healthcare environments. Such trials should incorporate long-term follow-up to evaluate the impact on colorectal cancer incidence, mortality, and cost-effectiveness.
Head-to-head comparative studies between different CAQ systems are essential to guide evidence-based selection and implementation. These studies should assess not only clinical performance but also usability, interoperability, workflow integration, and economic viability across varying healthcare settings.
Further research is also needed to determine the optimal integration of CAQ systems within existing endoscopy workflows to maximize benefits while minimizing disruption to established practices. This includes investigations into training requirements, change management strategies, and the long-term sustainability of quality improvements.
Finally, the development of standardized metrics and benchmarks for CAQ system performance is critical to enable comparison across platforms and facilitate meta-analyses. International collaboration in establishing these standards would promote global adoption and ensure consistent, high-quality performance evaluation across healthcare systems.
The emergence of CAQ systems in colonoscopy represents a promising advancement for enhancing the efficacy and consistency of procedures through standardized quality metrics, real-time feedback, and improved lesion detection. These systems demonstrate substantial potential to address existing gaps in quality control for screening colonoscopies and ultimately reduce colorectal cancer incidence. As the technology continues to evolve and the evidence base expands, CAQ systems may transform colonoscopy practice by enabling objective and standardized quality assessment, thereby improving patient outcomes and contributing to a reduction in the global burden of colorectal cancer. Nevertheless, widespread clinical adoption remains limited by several challenges, underscoring the need for careful implementation to ensure these tools complement—rather than complicate—clinical practice.
Fig. 1.
Schematic illustration of the integration of computer-aided quality control systems in colonoscopy. AQCS, automatic quality control system; BBPS, Boston bowel preparation scale.
ce-2025-309f1.jpg
Table 1.
Examples of CAQ systems in endoscopy
CAQ system Brief description/quality metrics measured
WISENSE12 Real-time monitoring of upper GI endoscopy that automatically tracks blind spots, inspection time, and completeness of photodocumentation to reduce missed areas and improve procedural quality.
ENDOANGEL (Shanghai Wision/ENDOANGEL)13 Real-time monitoring of withdrawal speed/time, blind-spot/coverage reminders, polyp detection overlays; studied in randomized and tandem trials showing increased ADR.
Automatic Quality Control System14 Measures intubation/withdrawal timing, withdrawal stability, evaluates bowel prep and flags polyps; shown to increase ADR in RCTs.
Endo.Adm15 An audit & feedback CAQ platform that automatically extracts key quality indicators (withdrawal time, cecal intubation, PDR/ADR, bowel prep) from videos/reports and issues performance reports.

CAQ, computer-aided quality control; GI, gastrointestinal; ADR, adenoma detection rate; RCT, randomized controlled trial; PDR, polyp detection rate.

  • 1. Hassan C, Spadaccini M, Mori Y, et al. Real-time computer-aided detection of colorectal neoplasia during colonoscopy: a systematic review and meta-analysis. Ann Intern Med 2023;176:1209–1220.ArticlePubMed
  • 2. Li JW, Chia T, Fock KM, et al. Artificial intelligence and polyp detection in colonoscopy: Use of a single neural network to achieve rapid polyp localization for clinical use. J Gastroenterol Hepatol 2021;36:3298–3307.ArticlePubMedPDF
  • 3. Li JW, Wu CC, Lee JW, et al. Real-world validation of a computer-aided diagnosis system for prediction of polyp histology in colonoscopy: a prospective multicenter study. Am J Gastroenterol 2023;118:1353–1364.ArticlePubMed
  • 4. Hassan C, Misawa M, Rizkala T, et al. Computer-aided diagnosis for leaving colorectal polyps in situ: a systematic review and meta-analysis. Ann Intern Med 2024;177:919–928.ArticlePubMedPDF
  • 5. Rizkala T, Hassan C, Mori Y, et al. Accuracy of computer-aided diagnosis in colonoscopy varies according to polyp location: a systematic review and meta-analysis. Clin Gastroenterol Hepatol 2025;23:531–541.ArticlePubMed
  • 6. Li JW, Wang LM, Ichimasa K, et al. Use of artificial intelligence in the management of T1 colorectal cancer: a new tool in the arsenal or is deep learning out of its depth? Clin Endosc 2024;57:24–35.ArticlePubMedPMCPDF
  • 7. Li JW, Wang LM, Ang TL. Artificial intelligence-assisted colonoscopy: a narrative review of current data and clinical applications. Singapore Med J 2022;63:118–124.ArticlePubMedPMC
  • 8. Li JW, Ang TL. Colonoscopy and artificial intelligence: Bridging the gap or a gap needing to be bridged? Artif Intell Gastrointest Endosc 2021;2:36–49.Article
  • 9. International Agency for Research on Cancer (IARC); World Health Organization (WHO). Colorectal cancer [Internet]. IARC, WHO; 2025 [cited 2025 Aug 1]. Available from: https://www.iarc.who.int/cancer-type/colorectal-cancer/
  • 10. Keswani RN, Crockett SD, Calderwood AH. AGA clinical practice update on strategies to improve quality of screening and surveillance colonoscopy: expert review. Gastroenterology 2021;161:701–711.ArticlePubMed
  • 11. Kaminski MF, Thomas-Gibson S, Bugajski M, et al. Performance measures for lower gastrointestinal endoscopy: a European Society of Gastrointestinal Endoscopy (ESGE) quality improvement initiative. United European Gastroenterol J 2017;5:309–334.ArticlePubMedPMCPDF
  • 12. Su JR, Li Z, Shao XJ, et al. Impact of a real-time automatic quality control system on colorectal polyp and adenoma detection: a prospective randomized controlled study (with videos). Gastrointest Endosc 2020;91:415–424.ArticlePubMed
  • 13. Schauer C, Chieng M, Wang M, et al. Artificial intelligence improves adenoma detection rate during colonoscopy. N Z Med J 2022;135:22–30.ArticlePubMed
  • 14. Lwin WP, Ichimasa K, Kudo SE, et al. Clinical significance of computer-aided quality assessment systems in colonoscopy: a comprehensive review. Clin Endosc 2025;58:638–645.ArticlePubMedPMCPDF
  • 15. Liu W, Wu Y, Yuan X, et al. Artificial intelligence-based assessments of colonoscopic withdrawal technique: a new method for measuring and enhancing the quality of fold examination. Endoscopy 2022;54:972–979.ArticlePubMedPMC
  • 16. Zhou J, Wu L, Wan X, et al. A novel artificial intelligence system for the assessment of bowel preparation (with video). Gastrointest Endosc 2020;91:428–435.ArticlePubMed
  • 17. Cold KM, Heen A, Vamadevan A, et al. Development and validation of the Open-Source Automatic Bowel Preparation Scale. Gastrointest Endosc 2025;101:1201–1210.ArticlePubMed
  • 18. Hsu WF, Chang WY, Kuo CY, et al. Effect of a novel artificial intelligence-based cecum recognition system on adenoma detection metrics in a screening colonoscopy setting. Gastrointest Endosc 2025;101:452–455.ArticlePubMed
  • 19. Chang YY, Li PC, Chang RF, et al. Development and validation of a deep learning-based algorithm for colonoscopy quality assessment. Surg Endosc 2022;36:6446–6455.ArticlePubMedPDF
  • 20. Filip D, Gao X, Angulo-Rodríguez L, et al. Colometer: a real-time quality feedback system for screening colonoscopy. World J Gastroenterol 2012;18:4270–4277.ArticlePubMedPMC
  • 21. Chen J, Wang G, Zhou J, et al. AI support for colonoscopy quality control using CNN and transformer architectures. BMC Gastroenterol 2024;24:257.ArticlePubMedPMCPDF
  • 22. Gong D, Wu L, Zhang J, et al. Detection of colorectal adenomas with a real-time computer-aided system (ENDOANGEL): a randomised controlled study. Lancet Gastroenterol Hepatol 2020;5:352–361.ArticlePubMed
  • 23. Thakkar S, Carleton NM, Rao B, et al. Use of artificial intelligence-based analytics from live colonoscopies to optimize the quality of the colonoscopy examination in real time: proof of concept. Gastroenterology 2020;158:1219–1221.ArticlePubMed
  • 24. Peterson E, May FP, Kachikian O, et al. Automated identification and assignment of colonoscopy surveillance recommendations for individuals with colorectal polyps. Gastrointest Endosc 2021;94:978–987.ArticlePubMed
  • 25. Wu L, Shi C, Li J, et al. Development and Evaluation of a Surveillance System for Follow-Up After Colorectal Polypectomy. JAMA Netw Open 2023;6:e2334822.ArticlePubMedPMC
  • 26. Koh FH, Li JW, Wong SH. Asia-Pacific consensus on the use of artificial intelligence in colorectal cancer screening and surveillance. iGIE 2025 Apr 16 [Epub]. https://doi.org/10.1016/j.igie.2025.04.001Article
  • 27. Okumura T, Kudo SE, Ide Y. Long-term impact of computer-aided adenoma detection: a prospective observational study. Endoscopy 2025 Sep 5 [Epub]. http://doi.org/10.1055/a-2661-2624Article
  • 28. Khouri A, Dickson C, Green A, et al. Effect of computer aided detection device on the adenoma detection rate and serrated detection rate among trainee fellows. JGH Open 2024;8:e70018.ArticlePubMedPMC
  • 29. Orzeszko Z, Gach T, Necka S, et al. The implementation of computer-aided detection in an initial endoscopy training improves the quality measures of trainees' future colonoscopies: a retrospective cohort study. Surg Endosc 2025;39:5276–5286.ArticlePubMedPMCPDF
  • 30. Yamaguchi D, Shimoda R, Miyahara K, et al. Impact of an artificial intelligence-aided endoscopic diagnosis system on improving endoscopy quality for trainees in colonoscopy: Prospective, randomized, multicenter study. Dig Endosc 2024;36:40–48.ArticlePubMedPMC
  • 31. Huang D, Shen J, Hong J, et al. Effect of artificial intelligence-aided colonoscopy for adenoma and polyp detection: a meta-analysis of randomized clinical trials. Int J Colorectal Dis 2022;37:495–506.ArticlePubMedPDF
  • 32. Soleymanjahi S, Huebner J, Elmansy L, et al. Artificial Intelligence-Assisted Colonoscopy for Polyp Detection : A Systematic Review and Meta-analysis. Ann Intern Med 2024;177:1652–1663.ArticlePubMedPDF
  • 33. Spada C, Salvi D, Ferrari C, et al. A comprehensive RCT in screening, surveillance, and diagnostic AI-assisted colonoscopies (ACCENDO-Colo study). Dig Liver Dis 2025;57:762–769.ArticlePubMed
  • 34. Hassan C, Spadaccini M, Iannone A, et al. Performance of artificial intelligence in colonoscopy for adenoma and polyp detection: a systematic review and meta-analysis. Gastrointest Endosc 2021;93:77–85.ArticlePubMed
  • 35. Seager A, Sharp L, Neilson LJ, et al. Polyp detection with colonoscopy assisted by the GI Genius artificial intelligence endoscopy module compared with standard colonoscopy in routine colonoscopy practice (COLO-DETECT): a multicentre, open-label, parallel-arm, pragmatic randomised controlled trial. Lancet Gastroenterol Hepatol 2024;9:911–923.ArticlePubMed
  • 36. Thomas J, Ravichandran R, Nag A, et al. Advancing colorectal cancer screening: a comprehensive systematic review of artificial intelligence (AI)-assisted versus routine colonoscopy. Cureus 2023;15:e45278.ArticlePubMedPMC
  • 37. Zhang Y, Zhang X, Wu Q, et al. Artificial intelligence-aided colonoscopy for polyp detection: a systematic review and meta-analysis of randomized clinical trials. J Laparoendosc Adv Surg Tech A 2021;31:1143–1149.Article
  • 38. Komanduri S, Dominitz JA, Rabeneck L, et al. AGA white paper: challenges and gaps in innovation for the performance of colonoscopy for screening and surveillance of colorectal cancer. Clin Gastroenterol Hepatol 2022;20:2198–2209.ArticlePubMed
  • 39. Zhang C, Yao L, Jiang R, et al. Assessment of the role of false-positive alerts in computer-aided polyp detection for assistance capabilities. J Gastroenterol Hepatol 2024;39:1623–1635.ArticlePubMed
  • 40. Nehme F, Coronel E, Barringer DA, et al. Performance and attitudes toward real-time computer-aided polyp detection during colonoscopy in a large tertiary referral center in the United States. Gastrointest Endosc 2023;98:100–109.ArticlePubMed
  • 41. Areia M, Mori Y, Correale L, et al. Cost-effectiveness of artificial intelligence for screening colonoscopy: a modelling study. Lancet Digit Health 2022;4:e436–e444.ArticlePubMed

Figure & Data

REFERENCES

    Citations

    Citations to this article as recorded by  
    • Artificial Intelligence in Colonoscopy Surveillance for Lynch Syndrome: Emerging Evidence, Lessons Learned From Average‐Risk Populations, and Future Directions
      Robert Hüneburg, Querijn N. E. van Bokhorst, Evelien Dekker, Jacob Nattermann
      International Journal of Cancer.2026;[Epub]     CrossRef

    • PubReader PubReader
    • ePub LinkePub Link
    • Cite
      CITE
      export Copy Download
      Close
      Download Citation
      Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.
      Format:
      • RIS — For EndNote, ProCite, RefWorks, and most other reference management software
      • BibTeX — For JabRef, BibDesk, and other BibTeX-specific software
      Include:
      • Citation for the content below
      Computer-aided quality control in colonoscopy: clinical applications and limitations
      Close
    • XML DownloadXML Download
    Figure
    • 0
    Related articles
    Computer-aided quality control in colonoscopy: clinical applications and limitations
    Image
    Fig. 1. Schematic illustration of the integration of computer-aided quality control systems in colonoscopy. AQCS, automatic quality control system; BBPS, Boston bowel preparation scale.

    Fig. 1.

    Computer-aided quality control in colonoscopy: clinical applications and limitations
    CAQ system Brief description/quality metrics measured
    WISENSE12 Real-time monitoring of upper GI endoscopy that automatically tracks blind spots, inspection time, and completeness of photodocumentation to reduce missed areas and improve procedural quality.
    ENDOANGEL (Shanghai Wision/ENDOANGEL)13 Real-time monitoring of withdrawal speed/time, blind-spot/coverage reminders, polyp detection overlays; studied in randomized and tandem trials showing increased ADR.
    Automatic Quality Control System14 Measures intubation/withdrawal timing, withdrawal stability, evaluates bowel prep and flags polyps; shown to increase ADR in RCTs.
    Endo.Adm15 An audit & feedback CAQ platform that automatically extracts key quality indicators (withdrawal time, cecal intubation, PDR/ADR, bowel prep) from videos/reports and issues performance reports.
    Table 1. Examples of CAQ systems in endoscopy

    CAQ, computer-aided quality control; GI, gastrointestinal; ADR, adenoma detection rate; RCT, randomized controlled trial; PDR, polyp detection rate.

    Table 1.

    Clin Endosc : Clinical Endoscopy Twitter Facebook
    © Korean Society of Gastrointestinal Endoscopy.
    Close layer
    TOP