To examine barriers and enabling strategies for the deployment of medical AI in low-resource settings (LRS), particularly in low- and middle-income countries (LMICs), emphasizing both challenges and solutions.
Approach:
Key Findings:
AI deployment in LRS is hindered by unreliable infrastructure, fragmented health data, and limited local skills.
Enabling strategies include investments in resilient digital infrastructure, interoperable data standards, and continuous capacity-building programs.
Human-centered values such as transparency, accountability, and equity are essential for sustainable AI integration, reflecting the thematic analysis.
Interpretation:
The review highlights that successful AI implementation in LRS relies more on systemic support and governance than on advanced technology alone, underscoring the need for robust governance frameworks.
Limitations:
Inclusion limited to English-language studies, which may affect the generalizability of findings.
Heterogeneity of studies prevented quantitative synthesis, limiting the ability to draw broader conclusions.
Conclusion:
Embedding human-centered values throughout the AI lifecycle is crucial for equitable and sustainable deployment of medical AI in LMICs, aligning with WHO and UNESCO AI ethics frameworks.
