CaseMatch

Inspiration

We wanted to explore how technology could make healthcare information easier to access, because healthcare professionals often have to search through large amounts of medical documentation to find relevant past cases, which can be time-consuming. This inspired us to create CaseMatch, a platform designed to make finding similar historical cases faster and more convenient.

What it does

CaseMatch is a clinical case search platform that helps healthcare professionals find past patient cases based on symptoms and other clinical information. Users can enter a search query and view related cases ranked by relevance. Each result provides a summary of the case, including general patient information and symptoms, but no personal patient data, making it easier to compare cases and identify useful information. Our overall goal is to make existing clinical information more accessible and easier to navigate.

How we built it

We built CaseMatch using React, TypeScript, Vite, and Tailwind CSS for the frontend, allowing us to create a responsive dashboard and search interface. For the backend, we used Python and FastAPI to manage requests and connect the application to a PostgreSQL database. We implemented semantic search using vector and matrix operations to calculate similarities between clinical cases and rank the results. We also designed a privacy-screening process using Presidio and spaCy to help protect sensitive patient information.

Challenges we ran into

One of our biggest challenges was figuring out how to organize clinical information in a way that was easy to navigate. We wanted search results to be detailed enough to be useful without overwhelming users. Another challenge was connecting the frontend and backend while making sure the search functionality worked as expected, especially within the 36-hour time frame of the hackathon.

Accomplishments that we're proud of

We are proud that we were able to bring database management and web development to create a platform that addresses a real-world problem, because we believe this tool could be truly useful for actual healthcare professionals. We especially liked developing a search system that goes beyond exact keyword matching and designing an interface that makes clinical information easy to navigate, and are proud that we could make patient privacy an important part of our project's design.

What we learned

Building CaseMatch gave us valuable experience with full-stack development, databases, and search functionality. We learned more about how different technologies work together and how important it is to design applications around the needs of their users. We also strengthened our teamwork and problem-solving skills while working under this tight deadline.

What's next for CaseMatch

In the future, we would like to expand our database of historical cases, improve search accuracy, and introduce more advanced search features. We would also like to strengthen patient privacy and security measures and eventually explore integration with existing hospital record systems. Our long-term goal is to make CaseMatch a genuine practical resource for healthcare professionals.