FUZZY COMPUTATIONAL MODELS TO EVALUATE THE EFFECTS OF AIR POLLUTION ON CHILDREN

Area of Science:

• Environmental Health
• Computational Modeling
• Pediatrics

Background:

• Air pollution and climatic factors are associated with increased respiratory conditions in children.
• Accurate prediction models are needed for effective public health management and resource allocation.

Purpose of the Study:

• To develop a fuzzy computational model to estimate pediatric hospitalizations due to respiratory conditions.
• To assess the association between specific air pollutants (PM10, NO2) and climatic factors (wind speed, temperature) with hospitalization rates in children up to 10 years old.

Main Methods:

• A fuzzy logic model (Mamdani's method) was constructed with 4 inputs and 16 rules.
• Hospitalization data (2011-2013) from DATASUS and environmental data from Cetesb were utilized.
• The model estimated the association between pollutants, climatic factors, and the number of hospitalizations.

Main Results:

• A total of 1,161 hospitalizations were recorded.
• Mean pollutant levels were 36 µg/m³ for PM10 and 51 µg/m³ for NO2.
• The model showed high accuracy, with the Receiver Operating Characteristic (ROC) curve indicating 96.7% accuracy for NO2 and 90.4% for PM10 on the same day of exposure.

Conclusions:

• The fuzzy computational model effectively predicted childhood respiratory hospitalizations.
• The model can serve as a valuable tool for hospital management in the studied region.
• This approach highlights the impact of environmental factors on pediatric respiratory health.