Enhancing detection rates, reducing failure rates in cancer screening

In a recent study published in the journal EClinical Medicineresearchers in China conducted a systematic review followed by a meta-analysis of studies describing the use of artificial intelligence (AI)-based methods to detect colorectal neoplasia during colonoscopies to understand its success in enhancing detection rates adenoma and in reducing adenoma loss rates.

Study: Artificial intelligence for colorectal neoplasia detection during colonoscopy: a systematic review and meta-analysis of randomized clinical trialssmall. Image credit: Peter Porrini / Shutterstock

Record

Colorectal cancer is one of the three most common cancers worldwide and contributes substantially to cancer-related mortality rates. Early detection of adenomatous polyps through colonoscopies and their removal is one of the main methods of reducing the incidence of colon cancer. While a 1% increase in adenoma detection rate is often associated with a 3% lower risk of colorectal cancer, variation among endoscopy services also results in an adenoma miss rate of approximately 27% due to cognitive or technical limitations.

Artificial intelligence-based methods have been extensively explored in recent years to standardize polyp detection during colonoscopies to circumvent human error. However, in addition to inconsistent results from studies examining the use of AI-based adenoma detection tools, there are also concerns about polyp overdiagnosis, leading to undue patient burden and cost. In addition, potential problems with endoscopy training and endoscopist distraction also present challenges.

About the study

In the present study, researchers conducted an extensive search for randomized controlled trials evaluating the advantages and disadvantages of using AI-based systems to detect adenomas and comparing them with standard colonoscopy-based detection methods. This comprehensive review and subsequent meta-analysis aimed to improve our understanding of AI-based detection methods for colorectal neoplasia.

Studies were included in the systematic review if the colonoscopy performed on enrolled participants was for primary screening, symptoms or surveillance, and studies compared AI-based colonoscopy methods with conventional colonoscopy methods. In addition, only randomized controlled trials that reported results relevant to this study were included. Studies involving patients with hereditary polyposis syndromes or inflammatory bowel disease were excluded.

Outcomes of interest were adenoma detection rate, adenoma loss rate, and adenomas detected at each colonoscopy. These primary outcomes were also stratified by morphology, pathology, location, and size. Secondary outcomes of interest were polyp detection rate, polyp miss rate, procedure time, false alarms, adverse events, and number of polyps detected at each colonoscopy.

Data extracted from studies included patient characteristics, study, intervention, polyps and adenomas detected, and primary and secondary outcomes. For consecutive trials, only the first colonoscopy data were used for the meta-analysis to prevent carryover effect. Heterogeneity between studies was quantified using the prediction interval, and subgroup and meta-regression analyzes were conducted to understand potential sources of heterogeneity.

Results

The results showed that the use of AI-based colonoscopy methods led to a significant enhancement of colon neoplasia detection and significantly reduced adenoma miss rate and polyp miss rate. Studies using AI-enabled colonoscopy also reported a significant increase in polyp detection rates and adenoma detection rates and the number of adenomas and polyps detected during each colonoscopy.

The polyp miss rate from AI-based colonoscopy methods was 52.5% lower, while the polyp detection rate was found to be 23.8% higher. Compared with conventional colonoscopy methods, the number of polyps detected per colonoscopy was 0.271 higher. However, the studies showed considerable heterogeneity in the outcomes associated with polyp detection.

Similarly, adenoma detection rate and adenoma miss rate showed a 24.2% increase and a 50.5% decrease, respectively, when AI-based colonoscopy methods were used. Additionally, approximately 0.202 more adenomas were detected per colonoscopy using AI-enabled adenoma detection methods. However, similar to polyp detection results, results from randomized controlled trials also showed considerable heterogeneity in outcomes.

conclusions

Overall, the findings suggested that the use of AI-enabled colonoscopy could significantly improve the detection of adenomas and colorectal neoplasia. Furthermore, slight improvements in the quality of colonoscopies could translate into potential net gains in large-scale colon cancer screening programs while maintaining homogeneity and quality of colonoscopies.

The researchers also discussed the future research implications of these findings, including the need for longitudinal studies to confirm the effectiveness of AI-based colonoscopic adenoma detection methods in reducing colorectal cancer-related morbidity and mortality.