Inspiration

Clinical trial recruitment is slow, confusing, and often inaccessible. Patients struggle to understand eligibility criteria, while researchers face delays finding qualified participants.

I wanted to make clinical trials easier to access and understand for everyone.

What it does

TrialMatch AI helps patients discover relevant clinical trials and understand whether they may qualify.

Users input basic information, and the system:

Matches them to relevant clinical trials Evaluates eligibility criteria Explains results in clear, human-readable language

Instead of complex medical jargon, patients receive simple explanations of why they are likely or unlikely to qualify.

How I built it

I built TrialMatch AI as a lightweight AI-powered web application:

Frontend: Simple interface for patient input and results display Backend: Python + Flask API for processing and matching Matching Engine: Rule-based logic combined with AI-driven explanations Data Layer: Structured clinical trial data (mock dataset for demo) AI Integration: Generates plain-language explanations of eligibility

I focused on clarity, usability, and real-world applicability.

Challenges I ran into

Translating complex eligibility criteria into understandable explanations Structuring clinical trial data in a usable format Designing outputs that are both accurate and user-friendly Building a clean UI under tight hackathon time constraints

Accomplishments that I'm proud of

Built an end-to-end patient-to-trial matching system in under 24 hours Successfully simplified complex medical eligibility into plain language Created a functional demo with real-world relevance Addressed a major bottleneck in clinical research recruitment

What I learned

The importance of explainability in AI systems Challenges in working with healthcare-related data How UX design impacts accessibility in sensitive domains Rapid full-stack development under pressure

What's next for TrialMatch AI

Integrate real clinical trial APIs (e.g., ClinicalTrials.gov) Improve matching accuracy with more advanced models Add patient profiles and saved results Expand to researcher-facing tools for recruitment optimization

Built With

Share this project:

Updates