What it does
The program we made is a screening method that we have developed to sort through around 1.6 million measurements from 580 patients over 6 months, and identify which patients may have certain conditions based on their test results. The program serves as a first step in finding early indications of the onset of certain conditions.
How we built it
We built it with Python and accessed the database in the FHIR API. Then we attempted to extract measures that are relevant to the diagnosis of the conditions of interest and sorted the patients according to the classifications established in current scientific literature.
Challenges we ran into
A big problem we had was setting up the connection to the database initially and figuring out how to extract only the relevant information from a list with over a million and several million lines of code. We had to figure out how to limit the output of the program to only the information we needed and also increase the speed and efficiency by finding patterns to help our search strategy.
What we learned
What an API is, how to access and use the FHIR API from Python, and what JSON is.
Accomplishments that we are proud of
When we started, we had no idea what an API is or what to do with the information. As four people who didn't know each other before starting we were able to work together to produce a pretty comprehensive project. We are proud of everybody's drive and tenacity that has been shown throughout this weekend.
What's next for ALVS Tactio Health Challenge
In the future, we would love to learn how to incorporate the back-end with a visually appealing front-end. Right now the program requires a lot of manual steps to be done by the user. In the future we would also like to automate this process.