SecurePatient

Inspiration

Companies like ChatGPT hold all your conversations. SecurePatient doesn't. SecurePatient uses End-to-End Encryption that is patient-controlled with SHA-256 Hashing, Base64 Encoding, and AES-256 Equivalent. It also uses Zero-Knowledge Proofs which verifies data without exposing any content. We create cryptographic proofs for each encrypted record.

We also feature a granular permission system (Read, ReadWrite, and Emergency). Only authorized parties can decrypt patient data We use Solana distributed ledger for data persistence and we store data hashes on Solana blockchain.

What it does

This project is a multi-agent healthcare diagnosis system that uses specialized AI agents to analyze patient symptoms, medical history, and lab results to provide transparent diagnostic recommendations for underserved communities. The system integrates with Solana blockchain to securely store health record hashes, making healthcare more private and accessible.

How we built it

We built this using Streamlit for the web interface and Python to orchestrate seven specialized AI agents: Intake & Preprocessing, Symptom Analysis, Medical History Analysis, Diagnostic Reasoning, Risk Assessment, Treatment Recommendation, and Results Validation. It integrates with Solana blockchain for secure record storage and uses natural language processing to analyze patient input, creating a multi-agent workflow that mimics (and possibly beats) real healthcare decision-making processes.

Challenges we ran into

The main challenge was coordinating multiple agents to work together while ensuring each agent's analysis fed properly into the next stage of the diagnostic pipeline. We also faced difficulties with blockchain integration complexity and handling the fallback when external dependencies like Solana or Gemini AI weren't available.

Accomplishments that we're proud of

We successfully created a working multi-agent system that shows the capability of transparent AI decision-making in healthcare, with each agent logging its activities for full auditability. The integration of blockchain technology for secure health record storage shows how emerging tech can be applied to existing healthcare infrastructure.

What we learned

We learned how to design and implement complex multi-agent systems where different AI components specialize in specific healthcare tasks while maintaining transparency and traceability. We also gained (super) valuable experience in integrating blockchain technology with healthcare applications and understanding the importance of fallback mechanisms when building systems that depend on multiple external services.