AI Identifies High-Risk Days for Cardiovascular Emergencies, Elderly Most Affected by Air Pollution

A cross-disciplinary team from National Taiwan Normal University has developed an artificial intelligence model capable of predicting high-risk days for cardiovascular emergencies. The study, published in the international journal "GeoHealth," integrates 23 years of environmental monitoring data with over 5 million emergency cardiovascular cases from Taiwan.

The AI model identified air pollution factors, specifically nitrogen oxides (NOx), as more significant predictors of high-risk days than traditional meteorological factors. The research highlights that the elderly population is the most vulnerable to environmental changes affecting cardiovascular health.

"Environmental factors are an important and quantifiable source of health risks," said Chen Hsiang-han, the lead author and an assistant professor of computer science. "The biggest challenge in this research was integrating 23 years of meteorological, air quality, and health data, and identifying the truly predictive signals from 184 environmental features."

The study found significant differences in sensitivity to environmental changes across various demographics. Men experience about 1.7 times the rate of acute cardiovascular emergencies compared to women. Crucially, individuals aged 65 and older have 2.4 times the rate of acute cardiovascular emergencies compared to those aged 50-64, and more than 11 times the rate of those aged 30-49. This underscores the diminished capacity of older adults to regulate in the face of environmental stressors.

Further analysis revealed that high-risk groups identified through air pollution factors showed a clearer distinction between high-risk and low-risk days compared to groups identified solely by weather conditions. Nitrogen oxides (NOx, NO, and NO2), which are linked to traffic emissions, along with atmospheric diffusion conditions like wind speed, were identified as key environmental indicators influencing emergency risk. The researchers suggest that integrating real-time air quality monitoring, weather forecasts, and health risk models could lead to the development of regional cardiovascular disease early warning systems.

“

Environmental factors are an important and quantifiable source of health risks. The biggest challenge in this research was integrating 23 years of meteorological, air quality, and health data, and identifying the truly predictive signals from 184 environmental features.