LLM-Assisted Scoping Review of Artificial Intelligence in Brazilian Public Health: Lessons from Transfer and Federated Learning for Resource-Constrained Settings

TL;DR

This study explores how AI, specifically Transfer Learning and Federated Learning, can help improve public health in Brazil by addressing data scarcity and privacy issues in resource-limited settings.

Contribution

The paper introduces a scoping review of AI applications in Brazil, emphasizing Transfer Learning and Federated Learning as novel solutions for public health challenges in the Global South.

Findings

Transfer Learning and Federated Learning are scalable and feasible for collaborative model training under privacy constraints.

These AI techniques remain underutilized in mainstream Brazilian healthcare despite their low resource requirements and data sovereignty benefits.

The study highlights the potential of resource-aware AI to promote equitable innovation in health systems of the Global South.

Abstract

Public health relevance—How does this work relate to a public health issue?
Maps how artificial intelligence is currently applied within the Brazilian Unified Health System (SUS), identifying structural gaps between diagnostic innovation and health system integration.Examines Transfer Learning and Federated Learning as practical responses to public health challenges such as data scarcity, privacy protection, and infrastructure limitations in the Global South.

Maps how artificial intelligence is currently applied within the Brazilian Unified Health System (SUS), identifying structural gaps between diagnostic innovation and health system integration.

Examines Transfer Learning and Federated Learning as practical responses to public health challenges such as data scarcity, privacy protection, and infrastructure limitations in the Global South.

Public health significance—Why is this work of significance to public health?
Demonstrates that resource-aware AI architectures can enable equitable innovation in large universal health systems without reliance on centralized data extraction or high-cost infrastructure.Provides empirical evidence that data sovereignty and cooperative AI development are achievable within real-world public health settings through decentralized and adaptive methodologies.

Demonstrates that resource-aware AI architectures can enable equitable innovation in large universal health systems without reliance on centralized data extraction or high-cost infrastructure.

Provides empirical evidence that data sovereignty and cooperative AI development are achievable within real-world public health settings through decentralized and adaptive methodologies.

Public health implications—What are the key implications or messages for practitioners, policy makers and/or researchers in public health?
For policymakers, the findings support Federated Learning as a governance-aligned strategy consistent with data protection laws (e.g., LGPD) and national digital health sovereignty.For researchers and practitioners, the review highlights Transfer Learning and Federated Learning as scalable, low-resource pathways to deploy clinically relevant AI tools in underserved and heterogeneous health systems.

For policymakers, the findings support Federated Learning as a governance-aligned strategy consistent with data protection laws (e.g., LGPD) and national digital health sovereignty.

For researchers and practitioners, the review highlights Transfer Learning and Federated Learning as scalable, low-resource pathways to deploy clinically relevant AI tools in underserved and heterogeneous health systems.

Artificial intelligence (AI) has become a strategic technology for global health, with increasing relevance amid the climate emergency and persistent digital inequalities. This study examines how AI has been applied in Brazilian healthcare through a scoping review with an in-depth methodological synthesis, focusing on Transfer Learning (TL) and Federated Learning (FL) as approaches to address data scarcity, privacy, and technological dependence. We searched PubMed, SciELO, and the CNPq Theses and Dissertations Repository for peer-reviewed studies on AI applications in Brazil, screened titles using AI-assisted tools with manual validation, and analyzed thematic patterns across methodological and infrastructural dimensions. Among 349 studies retrieved, six explicitly used TL or FL. These techniques were frequently implemented through multi-country research consortia, demonstrating scalability and feasibility for collaborative model training under privacy constraints. However, they remain marginal in mainstream practice despite their ability to deploy AI solutions with limited computational resources while preserving data sovereignty. The findings indicate an emerging yet uneven integration of resource-aware AI in Brazil, underscoring its potential to advance equitable innovation and digital autonomy in health systems of the Global South.

Figures