AI-Enhanced Cancer Surveillance in Lower Middle Income Countries: A Meta-Analysis of Effectiveness

Cancer remains a critical public health challenge in low- and middle-income countries, where limited resources and infrastructure often result in late-stage diagnoses and poor outcomes. Artificial intelligence (AI) offers transformative potential to enhance cancer surveillance in these settings. This meta-analysis synthesizes evidence from five studies evaluating AI applications in cervical, oral, urological, gastrointestinal, and thoracic cancer surveillance, assessing their effectiveness in lower middle income countries. Using a random-effects model, we report a pooled sensitivity of 88.5% (95% CI 83.2–92.6) and specificity of 84.3% (95% CI 78.9–88.7) across diagnostic or imaging cases, highlighting AI’s capacity to improve detection accuracy. Portable tools, such as smartphone-based oral cancer screening (sensitivity 96.7%) and enhanced visual assessment for cervical cancer (sensitivity 75.0%), demonstrate particular promise for resource-constrained environments. Policy implications include integrating AI into public health systems, enabling task-shifting to non-specialists, and addressing ethical concerns like algorithmic bias. Challenges, such as infrastructure limitations and costs, require tailored strategies, including offline-capable AI and public-private partnerships. Future research should focus on localized evidence, cost-effectiveness, and longitudinal impact to ensure equitable, sustainable AI deployment, ultimately strengthening health systems and advancing health equity in Lower Middle Income Countries.