Inspiration
Despite just being highschool students, we wanted to compete in this hackathon because we saw a growing need in America for cheaper prescription prices. Especially for the 30 million people who don't have insurance, rising drug prices mean that they cannot afford vital drugs. What's more, they can't even change their prescriptions(if one drug becomes cheaper than the one they are currently prescribed) because it would cost hundreds of dollars for them to go back to the doctor.
What it does
MediVision allows doctors to see the prices of certain drugs over time, and it will predict what the price of this drug will be 5 years from now. Therefore, when they prescribe drugs to the uninsured patients who come to them, they can give them the prescriptions that will cost them the least amount of money over time. We built an AI model in order to do so, training on the prices of thousands of drugs over time. We implement our own version sklearn's BaseEstimator, feeding in our data, measuring loss, and adjusting the model to fit the data. Finally, we compiled the results into an easily accessible application using Bokeh(to display our graph and build the UI).
What's next for MediVision
We plan on growing MediVision by adding more features(allowing doctors to compare multiple graphs of drugs at once), marketing it to doctors, and editing the application to include things like drug names instead of simply drug IDs.
Built With
- bokeh
- python
- sklearn
- visualization
Log in or sign up for Devpost to join the conversation.