Lung cancer is the leading cause of cancer mortality globally, despite the advancements in screening and management. Survival rates for lung cancer remain suboptimal, largely due to late-stage diagnoses and tumor heterogeneity. Recent advancements in artificial intelligence and radiomics provide a promising outlook for lung cancer screening, diagnosis, personalized treatment, and prognosis. These advances use large-scale clinical and imaging datasets that help identify patterns and predictive features that may be missed by human interpretation. Artificial intelligence tools hold the potential to take clinical decision-making to another level, thus improving patient outcomes. This review summarizes current evidence on the applications, challenges, and future directions of artificial intelligence (AI) in lung cancer care, with an emphasis on early diagnosis and personalized treatment. We examine recent developments in AI-driven approaches, including machine learning and deep neural networks, applied to imaging (radiomics), histopathology, biomarker analysis, and multi-omic data integration. AI-based models demonstrate promising performance in early detection, risk stratification, molecular profiling (e.g., programmed death-ligand 1 (PD-L1) and epidermal growth factor receptor (EGFR) status), and outcome prediction. These tools may enhance diagnostic accuracy, optimize therapeutic decisions, and ultimately improve patient outcomes. However, significant challenges remain, including model heterogeneity, limited external validation, generalizability issues, and ethical concerns related to transparency and clinical accountability. AI holds transformative potential for lung cancer care but requires further validation, standardization, and integration into clinical workflows. Multicenter collaborations, regulatory frameworks, and explainable AI models will be essential for successful clinical adoption.
Ayasa et al. (2025) studied this question.