Clinical Utility of Artificial Intelligence in Personalized Cancer Care: A Systematic Review of Evidence from Randomized Trials and Observational Studies

Artificial intelligence (AI) has revolutionized personalized cancer care by enabling data-driven prediction, diagnosis, and treatment optimization. This systematic review evaluates the clinical utility of AI in oncology based on evidence from ۱۲ high-quality studies, including ۵ randomized controlled trials and ۷ observational cohorts published between ۲۰۱۹ and ۲۰۲۴, encompassing over ۱۳,۰۰۰ patients with diverse malignancies. AI applications cover prediction of treatment response and toxicity, pathology and genomics-based tumor subtyping, clinical decision support, real-time therapy planning, and prognostic modeling using multi-omics data. Reported model performances showed AUCs ranging from ۰.۷۸ to ۰.۹۸ and C-indices between ۰.۶۸ and ۰.۹۲. AI-assisted interventions demonstrated improved clinical outcomes, such as longer progression-free survival and reduced toxicity rates. Despite these advances, challenges remain, including algorithmic bias, interpretability issues, and integration with electronic health records. In conclusion, AI holds substantial promise to enhance personalized oncology through improved prediction, diagnosis, and therapeutic decision-making. Continued multi-institutional validation, explainability, and clinician training are essential for safe and effective clinical implementation.

کلیدواژه ها:

Artificial intelligence ، Personalized cancer care ، Predictive modeling ، Clinical decision support ، Multi-omics integration