AI-powered disease characterization in colorectal and inflammatory bowel disorders

Colon diseases like colorectal cancer, Crohn’s disease, and ulcerative colitis are hard to diagnose and manage. Accurate early diagnosis lowers complications and death rates. Artificial intelligence, especially deep learning and computer vision, helps in detecting, segmenting, and classifying these diseases through medical imaging. Studies from 2015 to 2025 were reviewed using PubMed, IEEE Xplore, Web of Science, Scopus, and ScienceDirect. The search combined artificial intelligence methods with colon diseases and imaging techniques like colonoscopy, CT, MRI, and histology. Only peer-reviewed studies focused on AI-based diagnostic tools were included. CNNs, hybrid U-Net models, and CADe/CADx systems demonstrated high accuracy across diagnostic tasks. Public datasets like HyperKvasir and REAL-Colon, along with preprocessing techniques such as PCA and data augmentation, enhanced model performance. Integration of imaging with genomic and proteomic data supports personalized diagnostics. A conceptual hybrid AI framework is proposed to combine image-based analysis with clinical rules. AI improves diagnostic accuracy and efficiency in colon disease assessment. Future research should address explainability, generalizability, federated learning, and clinical validation for widespread adoption.