File size: 2,417 Bytes
7915030 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 | ---
title: TrialPath
emoji: 🏥
sdk: docker
app_port: 7860
colorFrom: blue
colorTo: purple
---
# TrialPath 🏥
AI-powered clinical trial matching system for NSCLC (Non-Small Cell Lung Cancer) patients.
## Overview
TrialPath helps patients understand which clinical trials they may qualify for by:
- Extracting patient information from medical documents using AI
- Matching patients to relevant clinical trials from ClinicalTrials.gov
- Providing gap analysis to transform "rejection" into actionable next steps
- Offering a clear patient journey with 5 guided steps
## Features
✅ **Multi-modal Document Upload** - Upload medical records, pathology reports, and lab results (PDF, images)
✅ **AI-Powered Extraction** - MedGemma 4B extracts structured patient profiles from unstructured documents
✅ **Smart Trial Matching** - Searches ClinicalTrials.gov with semantic understanding of eligibility criteria
✅ **Gap Analysis** - Identifies what's missing for trial eligibility and suggests next steps
✅ **Privacy-First** - No data storage, all processing in-session
## Tech Stack
- **Frontend**: Streamlit (Python)
- **AI Models**:
- Google Gemini 3 Pro (orchestration & planning)
- MedGemma 4B (medical document extraction)
- **Data Source**: ClinicalTrials.gov API v2
- **Workflow Engine**: Parlant (agentic framework)
## Current Status
🚧 **Proof of Concept** - Models and UI are functional with mock data. Live AI integrations in progress.
- ✅ UI: 5-page patient journey implemented
- ✅ Data Models: 5 Pydantic v2 contracts (PatientProfile, SearchAnchors, TrialCandidate, etc.)
- ✅ Services: MedGemma, Gemini, ClinicalTrials API clients ready
- 🚧 Agent: Parlant journey orchestration pending
- 🎯 Scope: NSCLC only, synthetic patients (no real PHI)
## Demo Mode
The app runs in demo mode by default with synthetic patient data. To enable full AI features:
1. Set `GEMINI_API_KEY` in Hugging Face Space secrets
2. (Optional) Set `MEDGEMMA_ENDPOINT_URL` for MedGemma extraction
3. (Optional) Set `HF_TOKEN` for Hugging Face authentication
## Local Development
```bash
# Install dependencies
pip install -r requirements.txt
# Run the app
streamlit run streamlit_app.py
```
## License
MIT License - See LICENSE file for details
## Contact
For questions or feedback, please open an issue on GitHub.
---
Built with ❤️ for patients navigating clinical trial enrollment
|