Supporting decision-making in the clinical management of dengue using predictive and prescriptive modeling

Abstract

Dengue is a vector-borne tropical disease that is widely distributed throughout the world. The clinical management of dengue consists of diagnosis and treatment of the disease. The development of approaches to aid decision-making in this disease is necessary to reduce morbidity and mortality rates. Despite the existence of guidelines for clinical management, the diagnosis and treatment of dengue remain a challenge. To address this problem, we aimed to develop methodologies, models, and approaches to support decision-making regarding the clinical management of this infection. We developed several works to meet the proposed objectives. First, we reviewed the latest trends in dengue modeling using machine learning (ML) techniques. Then, we proposed a decision support system for dengue diagnosis using fuzzy cognitive maps (FCM). Subsequently, we developed an autonomous cycle of data analysis tasks (ACODAT) to support both diagnosis and treatment of the disease. Also, we presented a methodology based on FCM and optimization algorithms to generate prescriptive models in various science domains. Finally, we proposed three federated learning approaches to ensure the security and privacy of data related to the clinical management of dengue. These strategies were evaluated using diverse datasets with signs, symptoms, laboratory tests, and information related to disease treatment. The results showed the ability of the developed methodologies and models to predict the disease, classify patients according to severity, evaluate the behavior of severity-related variables, and recommend treatments based on World Health Organization (WHO) guidelines.