Researchers used AI techniques to evaluate the association between built environment features identified by AI models and coronary heart disease (CHD). They used CNNs, LMEM, and activation maps to predict health outcomes at the census tract level using Google Street View images. The study found that AI algorithms could potentially reduce CHD burden by designing future cities with lower risk factors. While DSE factors were more accurate predictors of CHD, GSV features could still provide valuable information. The findings highlight the potential of machine vision-enabled identification of urban features related to CHD, which could lead to targeted interventions in high-risk neighborhoods.
Source link
AI deciphers city designs that could cut heart disease rates
