Artificial Intelligence in Lung Cancer: From Early Detection to Personalized Therapy.

purpose of reviewThis review highlights recent advances in the use of artificial intelligence (AI) for lung cancer screening, risk stratification, prognosis, and treatment selection, with a focus on emerging multimodal machine learning approaches. RECENT

findingsAI models, particularly deep learning algorithms, have exhibited high accuracy in interpreting imaging and liquid biopsy data for the early detection of lung cancer. Multimodal approaches integrating imaging, clinical, and molecular features have improved lung nodule classification and treatment response prediction. In retrospective studies, AI can outperform subjective decision-making thresholds and aid in providing patient-centered care. However, real-world deployment remains limited, and early clinical evaluations show performance degradation outside of retrospective settings. AI is poised to improve lung cancer care through enhanced detection, individualized prognostication, and personalized treatment. Multimodal models are especially promising but face challenges in generalizability, interpretability, and implementation. Clinical decision support systems that incorporate provider input may offer a feasible path to clinical integration. Prospective validation studies, system integration, and equitable data access are necessary to ensure AI can safely and effectively enhance lung cancer care.