Inspiration

Our project was born from a deep concern for patient safety and a passion for demystifying complex medical information. Inspired by a patient safety workshop and real stories from the field, we recognized that many patients struggle to understand medical documents. The recent 7.9 earthquake in one of our teammate's hometowns, where limited access to immediate medical help cost lives, further motivated us. We realized that if patients could receive clear guidance from an AI—even before an ambulance or doctor arrives—it could save lives.

What It Does

MediCura provides three core features:

  • Analyze Function:
    Converts complex lab reports and doctor's notes into plain, understandable language. It explains medical terms and suggests preventive measures to help patients take charge of their future health.
  • Chat Function:
    Acts as a virtual health advisor, enabling patients to ask questions such as "Can I take these two pills together?" or "Which diagnosis seems more accurate?" This feature offers quick, low-cost suggestions for minor diagnoses, alleviating the burden of expensive consultations.
  • Emergency Call Function:
    Offers immediate, step-by-step guidance in emergency situations, providing both medical instructions and emotional support during crises, such as natural disasters or when emergency services are delayed.

How We Built It

We developed MediCura using React Native Expo for a cross-platform mobile application. Key technologies include:

  • OCR: To extract text from scanned medical documents.
  • OpenAI: For AI-driven data analysis and clear explanations of complex medical cases. Together, these components enable our app to transform raw medical data into actionable insights in real time.

Challenges We Ran Into

Building MediCura in a 24-hour hackathon was no small feat. Some of our biggest challenges included:

  • Code Consistency:
    We encountered an issue where the same codebase ran on one computer but failed to connect on another. After several hours of backtracking through Git commits and clearing caches, we identified the issue and resolved it.
  • Handling Large Files:
    Our initial approach to processing large PDF lab reports via direct API calls failed due to file size limitations. After trying multiple methods, we implemented OCR, which effectively solved the problem.

Accomplishments We're Proud Of

We are immensely proud of our teamwork and perseverance in overcoming these challenges. In just 24 hours, we built a project that integrates three major AI features—each designed to help patients in different scenarios. We believe MediCura is both impactful and useful, addressing critical needs in healthcare.

What We Learned

Through this intense hackathon, we learned:

  • The importance of collaboration and collective problem-solving.
  • Effective debugging strategies for React Native.
  • Valuable insights into patient safety and the crucial need for AI tools that simplify medical information for everyday users.

What's Next for MediCura

Moving forward, we plan to implement a personal history feature. This will store a specific user's health data and, using a Model Context Protocol (MCP) technique, allow our AI to summarize previous diagnoses and provide personalized recommendations. With this enhancement, MediCura will deliver even more accurate and tailored guidance for every patient.

Share this project:

Updates