Comprehensive Summary
This systematic review examines the role of artificial intelligence (AI), artificial neural networks (ANNs), and machine learning (ML) in the management of non-small cell lung cancer (NSCLC), with a focus on patient-reported outcomes, overall survival, and cost-effectiveness. The authors conducted a PRISMA-guided review of the literature and synthesized findings from ten eligible studies using a narrative Synthesis Without Meta-analysis approach due to substantial methodological heterogeneity. Across included studies, AI-based models were primarily applied to imaging analysis, prognostic stratification, and treatment optimization, often outperforming conventional statistical methods in predictive accuracy for survival-related outcomes. While retrospective evidence suggests AI improves diagnostic precision and prognostic modelling, the review highlights persistent challenges related to data heterogeneity, model generalizability, and limited transparency. Importantly, the authors note that existing studies largely focus on technical performance rather than patient-centered or economic outcomes.
Outcomes and Implications
This review underscores a critical gap between technological advancement and clinically meaningful evaluation in NSCLC care. Although AI-driven tools show promise in improving prognostic assessment and supporting treatment decision-making, there is currently no evaluable evidence demonstrating benefits in patient-reported outcomes or cost-effectiveness. This limits conclusions about whether AI meaningfully improves patient experience, quality of life, or healthcare value. Future prospective studies should integrate validated PROM instruments and formal economic analyses alongside clinical endpoints to better assess real-world impact. Without these measures, widespread clinical adoption of AI in NSCLC risks prioritizing algorithmic performance over patient-centered care and equity.