NYC Gazette

Big data and artificial intelligence are transforming health paradigms, from disease detection to pattern recognition, predictive outcomes, and accelerated response times.

A recent study by New York University researchers analyzed two million Google Street View images from New York City streets to evaluate the utility of this digital data in public health decision-making. The findings, published in the Proceedings of the National Academy of Sciences (PNAS), indicate that while street view images alone may lead to inaccuracies and misguided interventions, their potential increases when combined with other knowledge sources.

“There’s a lot of excitement around leveraging new data sources to gain a holistic view of health, including bringing in machine learning and data science methods to extract new insights,” said Rumi Chunara, associate professor at NYU School of Global Public Health and NYU Tandon School of Engineering, and the study’s senior author.

“Our study highlights the potential of digital data sources such as street view images in enhancing public health research while also pointing out the limitations of data and the complex dynamics between the environment, individual behavior, and health outcomes,” added Miao Zhang, a PhD student at NYU Tandon School of Engineering and the study’s first author.

Researchers have increasingly used street view images to link an area’s environment with mental health, infectious diseases, or obesity—tasks challenging to measure manually. “We know that a city’s built environment can shape our health,” said Chunara. “Some studies show that the availability of sidewalks correlates with lower obesity rates—but is that the whole story?”

Chunara, Zhang, and colleagues analyzed over two million Google Street View images from every New York City street using AI to assess sidewalk and crosswalk availability. They compared this information with localized data on obesity, diabetes, and physical activity from the Centers for Disease Control and Prevention (CDC) to see if built environments predicted health outcomes.

The analysis revealed neighborhoods with more crosswalks had lower rates of obesity and diabetes. However, no significant link was found between sidewalks and these health outcomes. “This may be because many sidewalks in New York City are in places people don’t use—along highways or bridges—so sidewalk density may not reflect neighborhood walkability as accurately as crosswalks,” explained Zhang.

The researchers identified issues with AI-generated labels for street view images; some locations were incorrectly labeled due to cars or shade obscuring sidewalks in photos. In some instances without sidewalks were labeled as having them.

While crosswalks were linked to lower rates of obesity and diabetes, further analysis indicated physical activity accounted for these decreases more than crosswalk presence alone. Increasing physical activity could result in significantly larger decreases in obesity and diabetes than installing more crosswalks would achieve.

“We saw that physical activity delivers the benefits associated with crosswalks,” said Zhang. The researchers conclude that public health decision-making should incorporate domain knowledge alongside new data sources for effective interventions.

“While growing amounts of digital data can be useful in informing decision-making," added Chunara,"our results show that simply using associations from new data sources may not lead to useful interventions or best resource allocation.” A nuanced approach combining big data with expertise is needed for optimal use.

Salman Rahman and Vishwali Mhasawade from NYU Tandon also contributed to this study supported by the National Science Foundation (award 1845487).