Mapeamento dos métodos diagnósticos baseados em inteligência artificial: uma abordagem inovativa aplicada à hanseníase no Brasil (2014-2024)

Abstract

Leprosy remains a global public health challenge, with approximately 200,000 new cases detected annually, and Brazil is among the most affected countries. Despite advances in disease control, factors such as limited access to diagnosis, reliance on subjective clinical exams, and difficulties in contact tracing result in significant underreporting. Artificial intelligence (AI) emerges as a promising alternative to improve diagnostic accuracy, reduce detection time, and expand access to early diagnosis. However, its implementation faces methodological, regulatory, and structural challenges, requiring greater integration between science, innovation, and public policies. Although AI is transforming the diagnosis of several dermatological diseases, its application in leprosy is still in its infancy. The literature lacks in-depth studies on its feasibility and implementation, making it difficult to formulate guidelines for its clinical and epidemiological application. This research aims to fill this gap by analyzing how AI can contribute to improving diagnostic accuracy, expanding access to reliable tests, and supporting strategic recommendations for its incorporation into public policies. To this end, a Framework with an Innovative Approach (FAI) is proposed to map advances in early diagnostic methods for leprosy using AI (2014-2024). The study integrates three complementary approaches: scientific network analysis, technological patent prospecting, and public policy evaluation. This framework allows us to understand the relationship between academic production, technological innovation, and policy formulation for leprosy, providing strategic recommendations based on evidence. The research used data triangulation, combining scientific network analysis (VOSviewer, Gephi), technological prospecting (ORBIT Intelligence), and public policy evaluation (VantagePoint), allowing us to map trends and provide strategic recommendations. Unlike previous studies, this approach allows us to integrate academic knowledge, technological innovation, and government guidelines into a single analytical framework. The results indicate fragmentation in scientific collaboration on leprosy, with a strong dependence on a few hubs, which limits the dissemination of knowledge. The technological prospecting revealed the absence of Brazilian patents in the area, evidencing the disconnect between research and innovation. In addition, political guidelines still do not fully incorporate the potential of AI, reinforcing the need for greater alignment between science, technology and public policies to ensure its effective application. The research highlights the urgency of strengthening the interaction between science, innovation and public policies to optimize the use of AI in leprosy diagnosis. The FAI provides viable strategic recommendations to overcome structural challenges and stimulate the formulation of guidelines that encourage technological development and the modernization of leprosy control strategies. In addition, the findings of this thesis can serve as a methodological model for the application of AI in other neglected diseases, expanding the impact of this research in the field of global public health.