Real-time computer aided colonoscopy versus standard colonoscopy for improving adenoma detection rate: A meta-analysis of randomized-controlled trials

Abstract

Background: Recent prospective randomized controlled trials have evaluated deep convolutional neural network (CNN) based computer aided detection (CADe) of lesions in real-time colonoscopy. We conducted this meta-analysis to compare the adenoma detection rate (ADR) of deep CNN based CADe assisted colonoscopy to standard colonoscopy (SC) from randomized controlled trials (RCTs). Methods: Multiple databases were searched (from inception to May 2020) and parallel RCTs that compared deep CNN based CADe assisted colonoscopy to SC were included for this analysis. Using Mantel-Haenzel (M-H) random effects model, pooled risk ratios (RR) and mean difference (MD) were calculated. In between study heterogeneity was assessed by I 2 % values. Outcomes assessed included other per patient adenoma parameters. Findings: Six RCTs were included in our final analysis that utilized deep CNN based CADe system in real-time colonoscopy. Total numbers of patients assessed were 4962 (2480 in CADe and 2482 in SC group). CADe based colonoscopy demonstrated statistically higher pooled ADR, RR=1.5 (95% CI 1.3À1.72), p<0.0001, I 2 =56%; and pooled PDR, RR=1.42 (95% CI 1.33À1.51), p<0.00001, I 2 =9%; when compared to SC. Per patient adenoma detection parameters were significantly better with CADe colonoscopy when compared to SC, with increased scope withdrawal time (mean difference = 0.38, 95% CI 0.05À0.72, p = 0.02). Interpretation: Based on our meta-analysis, deep CNN based CADe colonoscopy achieved significantly higher ADR metrics, albeit with increased scope withdrawal time when compared to SC.

Artificial intelligence (AI) real-time detection vs. routine colonoscopy for colorectal neoplasia: a meta-analysis and trial sequential analysis

“…AI has established its utility with numerous luminal and extra-luminal applications in gastroenterology. Recent pooled studies have established its superior detection capabilities [30][31][32][33][34], and our study echoes the superior ADR and PDR rates of AI systems. Additionally, we included all prior studies to date and completed an in-depth subgroup analysis to parse out information based on colonic segments, size, and shape not described in other pooled studies.…”

“…Statistical significance and direction of the findings remained unchanged for ADR, PDR, mean adenomas or polyps detected per procedure, and WTs that included biopsies (Supplementary figures [25][26][27][28][29][30][31][32]. Overall mean WTs that were insignificant progressed to significance when Repici et al was removed (SMD 0.43; CI: 0.03-0.84, p = 0.04), yet the direction of findings favored AI-assisted colonoscopies.…”

Effect of artificial intelligence-aided colonoscopy for adenoma and polyp detection: a meta-analysis of randomized clinical trials

“…In other words, the use of AI in real-time colonoscopy does not appear to significantly improve the detection of advanced adenomas. This may be due to advanced adenomas being identified by endoscopists with extreme caution, or possibly the limited AAPC training data available for learning the CADe algorithm [6]. Serrated lesions (sessile serrated lesions and traditional serrated lesions) are notorious for causing CRC through serrated lesions pathway, accounting for roughly 20% of all CRC [29,30].…”

“…The effective detection of precancerous lesions with colonoscopy depends on various factors, including bowel preparation, the experience of the endoscopist, colonoscopy indication, physician fatigue, and examination time [6,30]. BPPS is an effective and reliable bowel preparation scale that reflects the cleanliness of the colon during colonoscopy [36].…”

“…Adenoma detection rate (ADR) is a well-accepted quality indicator of screening colonoscopy and is defined as the proportion of individuals who had at least one adenoma detected while undergoing a complete colonoscopy [6,7]. Each 1% increase in the ADR was associated with a 3% decrease in the risk of CRC [8].…”