Published Paper Details:

Artificial intelligence (AI) is increasingly recognized for its potential to enhance cancer care, particularly amid challenges such as an ageing population, workforce shortages, and rapid technological advancements. While AI has shown promise in cancer diagnostics, its broader impact on the quality, efficiency, and equity of care throughout the cancer care continuum remains uncertain. This systematic review aims to evaluate the clinical readiness and practical deployability of AI applications in post-diagnostic cancer care through an analysis of prospective studies. A comprehensive search of PubMed and Web of Science (1 January 2013 to 1 May 2023) identified 15 studies that prospectively assessed AI interventions following cancer diagnosis. Each study was critically appraised using the Risk of Bias tools for randomized and non-randomized studies, along with implementation analyses focusing on time, cost, and resource requirements. Our findings indicate that most AI research in oncology remains in the experimental phase, lacking prospective clinical validation or real-world deployment. Many studies failed to demonstrate clinical validity or translate AI-generated efficacy into meaningful patient outcomes. Key limitations include insufficient standardization, poor interoperability across health systems, and a lack of implementation and equity-focused evaluations. To address the existing gaps--namely, low clinical evidence, limited outcome translation, and research ecosystem fragmentation--future efforts should prioritize multicenter, collaborative implementation studies that align with current methodological standards and are tailored to specific healthcare settings.