Environmental Factors Linked to Cardiorespiratory Diseases Identified through Interpretable Machine Learning

Strategic Insights -

Researchers from the University of Basilicata have made a groundbreaking discovery in understanding the relationship between environmental factors and cardiorespiratory diseases. By applying machine learning techniques to a dataset of environmental and health data from 2013 to 2023, the team was able to identify specific environmental variables that contribute to an increase in daily emergency room admissions for cardiorespiratory diseases. The study highlights the importance of integrating environmental surveillance with healthcare planning to prevent and prepare for public health crises.

Key Takeaways:

  • The research used machine learning techniques to analyze the relationship between emergency room admissions for cardiorespiratory diseases and environmental factors.
  • Four machine learning models were compared, with XGBoost showing the best predictive performance (R^2 = 0.901; MAE = 0.047).
  • The model identified high levels of carbon monoxide and relative humidity, low atmospheric pressure, and mild temperatures as associated with an increase in CRD cases.
  • The study used LIME to identify critical thresholds beyond which a significant increase in the risk of hospital admission was observed: CO 0.84 mg/m^3, P_atm 1006.81 hPa, Tavg 17.19 °C, and RH 70.33%.
  • The findings emphasize the potential of interpretable machine learning models as tools for both epidemiological analysis and prevention support.
  • The research covers eleven years (2013-2023) of meteorological conditions and atmospheric pollutants data collected through meteorological sensors and air quality monitoring stations.

Statistics:

  • 11 years of environmental and health data were analyzed.
  • 4 machine learning models were compared, with XGBoost showing the best predictive performance.
  • R^2 = 0.901 and MAE = 0.047 were achieved using XGBoost model.
  • Carbon monoxide values above 0.84 mg/m^3 increased the risk of CRD cases.
  • Atmospheric pressure below 1006.81 hPa increased the risk of CRD cases.
  • Temperatures above 17.19 °C increased the risk of CRD cases.
  • Relative humidity above 70.33% increased the risk of CRD cases.

Sources:

  • An Interpretable Machine Learning Framework for Analyzing the Interaction Between Cardiorespiratory Diseases and Meteo-Pollutant Sensor Data (2025,25(15):4864) - Sensors (http://www.mdpi.com/journal/sensors).
  • University of Basilicata Researchers Advance Knowledge in Environmental Factors (2025, Ecology, Environment & Conservation).