Empowering public health: Leveraging AI for early detection, treatment, and disease prevention in communities - A scoping review.

Abstract

Abstract: India's healthcare system faces substantial challenges, including a high burden of communicable and non-communicable diseases, limited access to healthcare in rural areas, and a shortage of skilled healthcare professionals. Artificial intelligence (AI) offers promising solutions to address these gaps by enhancing diagnostic accuracy, improving disease prediction, and optimizing treatment management. This scoping review examines AI's role in early detection, treatment, and disease prevention in community health settings. A comprehensive literature search was conducted in PubMed, Embase, Scopus, and Google Scholar from January 2013 to July 2024. Eligible studies focused on the application of AI in public health, emphasizing early detection, disease prevention, and treatment interventions. Data on AI models, health outcomes, and performance metrics were extracted and analyzed in line with PRISMA-ScR guidelines. Forty-eight studies were analyzed and categorized into diagnostic accuracy, disease prediction, treatment management, and clinical validation. AI-based tools, such as AIDMAN for malaria detection, demonstrated high diagnostic accuracy (95%) and AUC (0.96). Predictive models for chronic kidney disease (93% accuracy) and diabetes (91% accuracy) showed substantial promise. TB screening using AI-powered cough analysis achieved 86% accuracy. The studies also emphasized AI's role in managing chronic diseases, facilitating early interventions, and reducing healthcare burdens in resource-limited settings. AI has the potential to revolutionize healthcare delivery in India, particularly in underserved regions, by enhancing early detection and treatment. However, challenges related to data privacy, algorithmic bias, and infrastructure require attention. Continued research and policy development are essential to fully harness AI's capabilities in improving public health outcomes.