Researchers have utilized Google Street View to analyze various elements of the built environment, such as buildings, green spaces, pavements, and roads, and their impact on coronary artery disease in different neighborhoods. The study found that these factors could predict 63% of the variation in the risk of coronary heart disease from one area to another. Coronary heart disease, caused by a build-up of fatty substances in the coronary arteries, is a prevalent form of cardiovascular disease. By assessing these environmental risk factors, researchers hope to contribute to creating healthier towns and cities.
The study was led by Prof. Sadeer Al-Kindi and Prof. Sanjay Rajagopalan from University Hospitals Harrington Heart & Vascular Institute and Case Western Reserve University in Ohio, along with Dr. Zhuo Chen, a post-doctoral fellow in Prof. Rajagopalan’s laboratory. They utilized machine vision-based approaches to evaluate the links between the built environment and the prevalence of coronary heart disease in various US cities. By analyzing over half a million Google Street View images from cities like Detroit, Cleveland, and Denver, as well as data on rates of coronary heart disease in smaller areas known as ‘census tracts,’ researchers were able to make predictions using a convolutional neural network.
The research findings showed that the features of the built environment visible on Google Street View images could effectively predict the variation in coronary heart disease between different regions in US cities. Attention mapping was also used to highlight important regions in the images, providing insights into factors that may influence cardiovascular risk. For example, green spaces and walkable roads were associated with lower risk, while poorly paved roads were linked to higher risk. These findings could potentially guide heart-healthy urban planning efforts and address upcoming challenges like climate change and shifting demographics.
Prof. Al-Kindi emphasized the importance of using computer vision approaches to identify environmental factors influencing cardiovascular risk, particularly on a large scale. With a significant portion of the global population expected to reside in urban environments, understanding urban environments using computer vision approaches is crucial for designing heart-healthy cities. The study’s innovative use of AI-enhanced methods to analyze the physical environment and its association with cardiovascular health highlights the role of visual appearances in determining cardiovascular health in communities. Strategic priorities must be identified to improve cardiovascular health in vulnerable communities based on these findings.
In an accompanying editorial, Dr. Rohan Khera from Yale University School of Medicine highlighted the significance of the study in comprehensively evaluating the built environment. By leveraging Google Street View imagery, the study offers a unique perspective on how the visual appearance of neighborhoods can encode measures of cardiovascular health. This information can be used to prioritize interventions in communities that require additional support to improve cardiovascular health. The study’s innovative approach underscores the importance of utilizing AI and computer vision technologies to assess environmental risk factors and address disparities in health outcomes across different communities. It is crucial to leverage this information to enhance cardiovascular health initiatives and support vulnerable populations at risk of coronary heart disease.
