JMIR Publications

ABSTRACT

Background:

For researchers interested in neighborhood built environment characteristics, Google Street View images provide a unique data source to examine these characteristics without the need for time-consuming and expensive in-person audits. Furthermore, most neighborhood studies have typically used cross-sectional designs.

Objective:

This study examined longitudinal changes in neighborhood built environments, demographic shifts, and health outcomes in Washington, DC from 2014 to 2019 by leveraging Google Street View images and computer vision.

Methods:

For this study, 434,115 Google Street View images were systematically sampled at 100-meter intervals along primary and secondary road segments. Convolutional neural networks, a type of deep learning algorithm, were used to automatically extract built environment features from images and create census tract summaries of the neighborhood built environment. Multilevel mixed-effects linear models with random intercepts at the year and census tract levels assessed associations between built environment changes and health outcomes, controlling for covariates such as census tract median age, percent male, percent Latinx, percent Black, percent with a college degree, and percent owner-occupied housing.

Results:

Over the study period, gentrification trends included an increase in college-educated residents (16% to 41%) and a rise in median property values exceeding $200,000, along with significant growth in non-single-family homes and road infrastructure. However, neighborhoods with higher proportions of African American residents experienced reduced construction activity. Longitudinal analyses revealed that increased construction activity was associated with lower rates of obesity, diabetes, high cholesterol, and cancer, while growth in non-single-family housing correlated with reductions in obesity and diabetes.

Conclusions:

These findings highlight the complex interplay between urban development, demographic changes, and health, underscoring the need for future research to explore the broader impacts of neighborhood built environment changes on community composition and health outcomes. Street View imagery along with advances in computer vision can aid in the acceleration of neighborhood studies.