Android home page
Specific lesion page
Home screen of web client
We were inspired by the med-bays in the movie Elysium, where patients were diagnosed and treated in an instant.
What it does
Elysium allows users to quickly and quite accurately analyze the potential risk of lesions present on the skin, and communicate these findings directly to their healthcare professionals for input and easy appointment making.
How we built it
The mobile client is built using Android, the web client was built using Node.JS, and our custom machine learning model was a custom Microsoft Azure Custom Vision API.
Challenges we ran into
We had difficulties figuring out how to interact with the API on the Android app, and also using Firebase Firestore, which we were completely new to.
Accomplishments that we're proud of
Creating a functional platform which reached an accuracy of 78-80% on both Android and the web in 36 hours.
What we learned
How to use Azure's custom suite of APIs, how to use Firebase.
What's next for Elysium
Integrating our own model to introduce more granularity in the data.