Inspiration
Many of our loved ones suffer from rare diseases, and based on their experiences, getting answers took months, despite living in an age of information. We wanted to build a network that uses machine learning to extract health-related information in a secure fashion and quickly match users to related clinical trials and connect them with possible treatments. Additionally, we've wanted to work with Amazon Web Services, and their Comprehend Medical API was the perfect fit for our project.
What it does
A patient can simply copy and paste their chart, and the website will return matching clinical trials and the estimated match. Beyond the primary feature, we've implemented a clinical trial directory, where users can browse the trials that are recruiting participants.
How we built it
We created the frontend use Vite and React, and the backend connects to a MongoDB database through Express.js and Node.js. The backend routes were tested using Postman. To implement a matching function, we used string-similarity-js, which is a javascript library that calculates similarity based on the Sorensen-Dice coefficient.
Challenges we ran into
The hardest part was making the request to the MongoDB database while simultaneously making an API call to the Amazon AWS Comprehend Medical service.
Accomplishments that we're proud of
We successfully created a full-stack application that connected AWS and MongoDB, something we've never done before. We also learned about front-end design, and created an svg logo from scratch.
What we learned
We learned more about the services that AWS provides, specifically AWS Comprehend Medical. Some of our group had never used React/JS before, and this project allowed them to learn about frontend development and web design.
What's next for Sirona Clinical
We would love to continue implementing features that enable clinical trial organizers to publish trials and garner interest, which would populate our clinical trial database.
Log in or sign up for Devpost to join the conversation.