rtificial Intelligence in Oncology Clinical Decision Support: From Diagnosis to Precision Cancer Care

Abstract

Background: Clinical Decision Support Systems (CDSS) have become an integral component of modern oncology by assisting clinicians in diagnosis, treatment planning, prognostic assessment, therapeutic monitoring, and personalized patient management. The increasing complexity of cancer care, driven by rapidly expanding genomic information, advanced imaging modalities, biomarker discovery, immunotherapy, and precision medicine, has created an unprecedented demand for intelligent computational tools capable of integrating heterogeneous clinical data. Artificial intelligence (AI), particularly machine learning, deep learning, natural language processing, reinforcement learning, and multimodal foundation models, has significantly enhanced the capabilities of oncology CDSS by enabling accurate prediction, automated interpretation, risk stratification, and evidence-based clinical recommendations. AI-driven CDSS can integrate radiological imaging, digital pathology, genomic sequencing, electronic health records, laboratory investigations, wearable sensor data, and published clinical evidence to support multidisciplinary decision-making throughout the cancer care continuum. Recent advances have demonstrated substantial improvements in tumor detection, molecular subtype prediction, treatment selection, immunotherapy response prediction, toxicity assessment, survival estimation, and clinical workflow optimization. Nevertheless, important challenges remain regarding model interpretability, algorithmic bias, regulatory approval, cybersecurity, interoperability, and ethical implementation. This review discusses the evolution, computational architecture, major clinical applications, current limitations, and future perspectives of artificial intelligence-powered clinical decision support systems in oncology. It also highlights how intelligent CDSS are reshaping precision cancer care by improving diagnostic accuracy, therapeutic personalization, and clinical efficiency while supporting evidence-based oncology practice.