A neuro-fuzzy model to predict respiratory disease hospitalizations arising from the effects of traffic-related air pollution in São Paulo

The significant volume of vehicular traffic has been considered one of the main causes of air pollution due to the rapid growth of urbanization and motorization in the world. This trend has instigated efforts to search for sustainable solutions aimed not only at mitigating the deleterious consequences stemming from air pollution but also at implementing efficacious urban mobility strategies and policies. In this context, the present study endeavors to explore the modeling and predicting of hospitalizations and associated costs linked to respiratory diseases, influenced by vehicular pollutants within the urban milieu of São Paulo—a city renowned for harboring one of the largest vehicular fleets globally. Specifically, an adaptive neuro-fuzzy inference system (ANFIS) was developed based on pollutant data encompassing carbon monoxide (CO), Particulate matter with diameters less than 10 µm (PM₁₀), Particulate matter with diameters less than 2.5 µm (PM_2.5), nitrogen dioxide (NO₂), oone (O₃), and sulfur dioxide (SO₂), emitted within the city confines spanning the period from 2011 to 2019. The simulations conducted revealed that with knowledge of the monthly concentrations of the analyzed pollutants, it was feasible to forecast hospitalization rates and costs with an error lower than 6%. Additionally, scenarios illustrating the applicability of ANFIS in public health management and its contributions to the United Nations Sustainable Development Goals (SDGs) are presented and discussed.