Skip to content

Medical Intelligence AI

Quick Background for New Users

New to this case? Read the 📋 Medical Summary Portfolio for a comprehensive overview of diagnosis, treatment history, and current status.

Ask questions about your health records using natural language. The AI has access to all your blood tests, medical reports, treatments, and timeline data.

💡 How to Use

  • Ask Tab: Natural language questions (e.g., "What's my latest β-HCG value?")
  • Search Tab: Direct vector search for keywords or phrases
  • Stats Tab: Overview of all indexed medical records

🎯 Example Questions

  • "How many hyperthermia sessions do we have on record?"
  • "What was my β-HCG on November 17, 2025?"
  • "Show me all my chemotherapy cycles"
  • "Is there a correlation between hyperthermia and β-HCG?"
  • "What did the latest PET scan show?"
  • "What's my white blood cell trend?"

AI Interface

Features

🤖 AI-Powered Question Answering

  • Natural language understanding
  • Contextual responses with source citations
  • Multi-turn conversations
  • Direct data access via tool calling
  • Semantic search across all medical records
  • Filter by collection type (blood tests, reports, treatments)
  • Date range filtering
  • Similarity scoring

📊 Data Coverage

  • Blood Tests: All lab results with trends
  • Medical Reports: CT scans, histology, ultrasounds, etc.
  • Treatments: Chemotherapy, hyperthermia, surgery timeline
  • Medical Events: Consultations, scans, procedures

🔒 Privacy & Security

  • Runs locally on your machine
  • No external API calls for data storage
  • OpenAI API used only for LLM processing
  • Your medical data never leaves your control

Technical Details

Technology Stack: - Vector database (ChromaDB) for semantic search - OpenAI GPT-4 for natural language understanding - FastAPI backend with real-time responses - Tool calling for direct data access

Data Sources: - data/processed/blood_tests.csv - 20 β-HCG tests - data/processed/treatment_timeline.json - 97 treatment events - data/processed/medical_reports_catalog.json - 25+ medical reports - Full text extraction from all PDFs and documents

Tips for Best Results

  1. Be specific: "What was my β-HCG on July 15?" works better than "What's my β-HCG?"
  2. Ask follow-ups: The AI remembers conversation context
  3. Request comparisons: "Compare my WBC in June vs October"
  4. Ask for trends: "How has my platelet count changed?"
  5. Get summaries: "Summarize my treatment timeline"

Limitations

Important Disclaimer

This AI tool is for informational purposes only and does not provide medical advice. Always consult with your healthcare providers for:

  • Treatment decisions
  • Interpretation of test results
  • Medical diagnoses
  • Health concerns or symptoms

The AI works with the data available in the system. If recent records haven't been processed yet, they won't appear in responses.

Need Help?

  • Data not showing? Make sure PDFs have been processed and indexed
  • Unexpected answers? Try rephrasing your question more specifically
  • Technical issues? Check that the Docker containers are running

Last Updated: November 25, 2025