Diagnostic and Screening AI Tools in Brazil’s Resource-Limited Settings: Systematic Review

Comprehensive Summary

Mancini et al. review how artificial intelligence is applied in Brazilian health care, specifically in resource-limited settings. Mancini et al. searched six major databases for studies from 1993–2023, chose 25 relevant papers, and performed meta-analyses on sensitivity, specificity, and AUC using a random-effects model. The review found that most AI tools in Brazil were used for diagnosing or screening diseases, especially in ophthalmology and infectious disease. Overall, the AI tools showed good accuracy (81% sensitivity, 74% specificity, and 83% AUC). Mancini et al. highlight that AI has a strong potential to improve diagnosis and screening in Brazil but emphasize the need for more testing on different populations to confirm reliability. Mancini et al. also emphasize that limited technological infrastructure and financial constraints must be addressed to make AI tools practical in resource-limited health care settings.

Outcomes and Implications

This research is important because it shows how AI can improve health care in resource-limited health care settings, like Brazil, while pointing out what is needed to make these tools more effective and accessible in these settings. This work applies to medicine by showing that AI can support physicians in diagnosing and screening diseases more accurately and efficiently, especially in areas with limited medical resources. Its clinical relevance lies in its potential to improve patient outcomes and expand access to quality care across underserved regions in Brazil.