patientzero 0.1.0

A generic framework for LLM-driven patient simulations.

PatientZero

Explaining Health: How AI Explanation Styles and Interaction Modalities Affect User Comprehension of Medical Information

Overview

A simulation system that uses LLM-powered agents to simulate doctor-patient interactions around medical test results and evaluate patient comprehension.

Tech Stack

• Frontend: React + Vite + TypeScript, Tailwind CSS, shadcn/ui
• Backend: Python, FastAPI, SQLite
• LLM: Abstracted provider layer (Mock, OpenAI, Claude, Local)

Project Structure

PatientZero/
├── core/              # Domain logic (imported by the backend)
├── frontend/          # React app
├── backend/           # FastAPI HTTP layer
└── report.txt         # Research report

Prerequisites

• Python 3.12+
• Node.js 20+
• uv (Python package manager)

Quick Start

1. Clone the repo:

git clone https://github.com/tamteaa/PatientZero.git
cd PatientZero

2. Set up environment:

cp .env.example .env

3. Install Python deps and start the backend (from the repo root so core imports resolve):

uv sync
uv run uvicorn backend.api.main:app --reload

4. Start the frontend:

cd frontend
npm install
npm run dev

5. Open http://localhost:5173

License

MIT