ARTIFICIAL INTELLIGENCE APPLICATION IN COLONOSCOPY SCREENING: A LITERATURE REVIEW

Background: Colorectal cancer (CRC) is the third most commonly diagnosed cancer globally and remains a leading cause of cancer-related deaths. Despite the effectiveness of colonoscopy in reducing CRC incidence and mortality through adenoma removal, some polyps are frequently missed. Artificial intelligence (AI) has recently appeared as a promising tool to enhance detection rates during colonoscopy. Aim of the study: This study aims to compare AI-assisted colonoscopy with standard colonoscopy in terms of adenoma detection rate and polyp detection rate. The goal is to evaluate whether one approach is superior to the other. Material and methods: A systematic literature search of the PubMed database was performed for studies published between 2015 and 2025. Search terms included “artificial intelligence”, “machine learning”, “colonoscopy”, and “mass screening”. Only English-language studies directly comparing adenoma and polyp detection rates between AI and standard colonoscopy procedures were included. A total of 18 studies involving 12,000 patients met the inclusion criteria. Results: The AI group consistently demonstrated higher adenoma detection rates compared to the standard colonoscopy group, with 36.06% vs. 28.85%, respectively. Similarly, AI showcased greater polyp detection rates, detecting polyps in 41.05% of patients compared to 34.17% in the standard colonoscopy group. Advanced AI techniques reported the highest detection rates. AI showed enhanced performance in identifying diminutive lesions and polyps located in challenging regions. Importantly, AI did not significantly prolong withdrawal times. Conclusions: AI integration into colonoscopy improves adenoma and polyp detection rates across diverse patient populations and clinical settings. Even among experienced endoscopists, AI provides added diagnostic value. The findings highlight AI’s potential to enhance CRC screening, though further studies are needed to standardize AI tools, validate their efficacy in real-world settings, and assess long-term clinical outcomes.