In a review article recently published in the journal Annals of Internal Medicine, researchers comprehensively examined the effectiveness of artificial intelligence (AI)-assisted colonoscopy in improving the detection of colonic polyps and adenomas during routine screenings.
Study: Artificial Intelligence–Assisted Colonoscopy for Polyp Detection: A Systematic Review and Meta-analysis. Image Credit: Kzenon/Shutterstock.com
The study aimed to compare enhanced colonoscopy with a computer-aided detection (CADe) system against traditional methods, focusing on adenoma detection rate (ADR) and other relevant metrics. The study ultimately assessed whether AI technology could improve diagnostic outcomes in colorectal cancer screening.
AI in Medical Diagnosis
The integration of AI in medical diagnostics is revolutionizing many fields, including disease detection through imaging. In gastrointestinal health, colonoscopy is the gold standard for screening and detecting colorectal cancer. However, traditional methods are prone to error, which can lead to missed adenomas and polyps that may progress to cancer if not detected.
CADe systems are designed to address these limitations. In colonoscopy, CADe systems utilize advanced algorithms, mainly based on machine learning and various neural network architectures (such as convolutional neural networks), to support endoscopists in detecting polyps and other neoplastic lesions.
These systems analyze real-time video feeds during procedures, aiming to boost the ADR and lower the adenoma miss rate (AMR), both critical for preventing and diagnosing colorectal cancer. While initial studies showed promising results, the effectiveness of these systems in real-world settings is still under investigation.
About the Review
In this paper, the authors conducted a systematic review of 44 randomized clinical trials (RCTs), encompassing a total of 36,201 cases. They performed a thorough search across multiple databases, including the Cochrane Library, PubMed, and Scopus, to identify studies comparing CADe-enhanced colonoscopy with conventional techniques.
The primary outcomes measured were the average adenoma per colonoscopy (APC) and advanced colonic neoplasia (ACN) per colonoscopy. Secondary outcomes included the ADR, AMR, and ACN detection rate (ACN DR), while additional parameters such as withdrawal time and resection of non-neoplastic polyps (NNPs) were also evaluated.
The study used rigorous protocols for data extraction, aiming to balance outcomes like withdrawal time and resection of NNPs. Subgroup analyses examined different neural network architectures to assess performance variations, providing a comprehensive evaluation of CADe systems' effectiveness in clinical settings.
Key Findings and Insights
The meta-analysis revealed that CADe-enhanced colonoscopies significantly outperformed conventional methods in several key metrics. Specifically, the average ADR was notably higher in the CADe group, indicating a substantial improvement in the identification of adenomas.
The analysis indicated a rate ratio (RR) of 1.21 for ADR with CADe systems, highlighting the statistical significance of these findings. Furthermore, while the average advanced colorectal neoplasia per colonoscopy showed no significant difference between the two methods, the detection rate for advanced colorectal neoplasia was notably higher in the CADe group, suggesting its effectiveness in identifying serious lesions.
Additionally, the use of CADe systems resulted in the resection of an increased number of non-neoplastic polyps, along with a slight increase in total withdrawal time during procedures.
The study also acknowledged several limitations, including significant heterogeneity in the quality and sample sizes of the included studies. The inability to blind endoscopists to the intervention in some studies could potentially bias the results, underscoring the need for cautious interpretation of the findings.
Applications
The implications of this research extend beyond academic interest, holding practical significance for clinical practice and patient outcomes. The enhanced detection rates associated with CADe systems could lead to earlier identification and treatment of colorectal neoplasia, ultimately improving patients' prognosis and survival rates.
As the prevalence of colorectal cancer continues to rise globally, integrating AI technology into routine screening practices could alleviate some of the burdens on healthcare systems by reducing the number of missed adenomas.
The authors suggest that ongoing training and adaptation of endoscopists to utilize AI effectively can enhance procedural efficiency and accuracy. Future applications may include personalized screening protocols based on individual risk factors, further optimizing patient management.
Conclusion
In summary, this systematic review and meta-analysis indicated that CADe-enhanced colonoscopy significantly improves ADR compared to conventional methods, with minimal increases in procedure time.
The study highlighted the potential of AI in enhancing colorectal cancer screening, emphasizing the need for further research to address existing limitations and refine these systems for broader clinical application.
The integration of AI in colonoscopy represents a promising advancement in the fight against colorectal cancer, with the potential to improve patient outcomes and enhance the overall quality of care.
Journal Reference
Soleymanjahi, S., & et al. Artificial Intelligence–Assisted Colonoscopy for Polyp Detection: A Systematic Review and Meta-analysis. Annals of Internal Medicine, 2024. DOI: 10.7326/ANNALS-24-0098, https://www.acpjournals.org/doi/10.7326/ANNALS-24-00981
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.