Hospital wait times are staggeringly long. For the single mother who has to work during the day and take care of her kids at night, these wait times make quality care inaccessible. We want to change that, so we made a more intelligent queue system that prioritizes high risk patients, and increases access to care by notifying patients exactly when they should come in for immediate attention.

What it does

Patients with different combination of symptoms take different lengths of time to treat. Patients with coughs may take 5 minutes, while patients with fevers may take 10 minutes; patients with both symptoms may suggest a more severe diagnosis and take up to 90 minutes. To accurately predict wait times and address more severe patients earlier, we use multiple regression to gauge patient risk. We then schedule patients efficiently into doctors' schedule in order of risk.

How we built it

We used Flask as the web server hosted on Heroku. Frontend was built with pure HTML, CSS, and Javascript. Google Calendar APIs was used to automatically schedule appointments and determine physician availability. We also used R for our multiple regression model. The algorithm used in scheduling patients is the knapsack algorithm with greedy approach.

Challenges we ran into

We struggled with Heroku dependency issues as well as callback bugs with Google APIs.

Accomplishments that we're proud of

We finished a relatively delicate web app that can illustrate proof of concept.

What we learned

We learned the intricacies of using Flask as well as how to make our UI and security appropriate for the medical industry.

What's next for Qster

We plan on polishing infrastructure and showing companies what we can do with our new system!

Share this project: