ScholarWorks@중앙대학교

초록

The application of artificial intelligence (AI) and machine learning (ML) in colorectal cancer (CRC) risk prediction models has been increasingly used in clinical and research settings. AIbased models have been used to predict the risk of CRC and high-risk polyps by integrating diverse data, including clinical information, genetic factors, lifestyle factors, and imaging data. Various ML algorithms, such as random forest, artificial/deep neural networks, and support vector machines, are currently applied and often demonstrate a higher sensitivity and specificity than traditional statistical prediction models. In particular, AI has been shown to improve prediction accuracy by combining multimodal data sources, such as electronic health records, genomic information, and pathology images, thereby supporting personalized screening strategies. Furthermore, this study aimed to investigate the latest trends, performance, external validation results, integration into clinical workflow settings, and the future challenges of AI models for CRC risk prediction.

키워드