Diabetes is the sixth leading cause of death in the United States. There is a strong and well known relationship between obesity and diabetes. Obesity has an infectious quality so much so that epidemiological approaches have been used to study obesity. It's also been reported that social factors can contribute to obesity and diabetes. We believe that by studying social networks and the health status of those within any given individual's social network, we can produce a risk score that can help drive the conversation between clinicians and patients to improve pre-diabetic management.

What it does [theoretically]

  • User logs in to the application through their social network account
  • It then mines the relationship between user and other individuals (i.e. how many photos do they share together, or how often do they leave comments on each others' posts), and establishes a 'closeness' score for each individual and the user.
  • It also assesses the health status of the individuals via secondary means or by interfacing with their electronic health records, and categorizes them into four buckets - healthy, obese, pre-diabetic, and diabetic.
  • Based on the 'closeness score' and the severity of the individuals' condition, we evaluate the likelihood that the user will develop diabetes. This analysis may be supplemented with analysis of the electronic health record itself.
  • This webapp can be embedded into the electronic health record or simply shared in a clinicians' office to help the doctor and patient communicate with each other.

What it actually does

It produces a d3 force-directed graph of fabricated social network dataset with fabricated health status and a random risk score upon clicking the 'connect to social media' button.

How I built it

I used node.js, express.js, d3.js, jquery, forever.js, put on an EC2 AWS instance, hosted on a personal domain.

Challenges I ran into

I didn't have access to real social network data or individuals' health status. In fact, this app could only exist in what some would call a dystopian future where social media accounts can be easily linked to electronic health records. Alternatively, we could estimate an individuals' obesity status by scanning their images, but that's heading into socially/ethically muddy area as well. All that being said, I think the general concept of leveraging social networks to better understand social determinants of health has some merit.

Accomplishments that I'm proud of

I finished a project.

What I learned

Learning new skills within 24 hours is difficult. Finding a team with complimentary skills & commitment is difficult.

But Most of All

Thank you ArchHacks 2017 for the opportunity to meet new people and develop a cool thing :)

Share this project: