🩺 Tammeny — طمّني

AI-Powered Healthcare Communication

Bilingual health assistant for Arabic & English speakers.
Ask questions, upload medical files, analyze X-rays, and search medical sources — all in plain language.

---

Screenshots

Bilingual Chat

Ask any health question in English or Arabic. Tammeny responds in your language with warm, clear advice — and always reminds you to see a doctor.

---

Chest X-Ray Analysis — HIGH Severity

Upload a chest X-ray and get instant AI screening for 14 thoracic conditions. Each finding includes a confidence score and a plain-language explanation of what it means.

---

Chest X-Ray Analysis — LOW Severity

When findings are minor, Tammeny clearly reassures the patient — no panic, just the right next step.

---

Demo Video

chrome_kPLT1Zsh6A.mp4

---

Features

Feature	Model	Description
💬 Bilingual Chat	Mistral (small-latest)	Health Q&A in Arabic & English with conversation memory
📄 PDF Document Q&A	Mistral + FAISS	Upload a medical PDF and ask questions about it
🔬 Medical Image Analysis	Qwen2.5-VL-7B	Plain-language explanation of any medical image
🩻 Chest X-Ray Analysis	DenseNet-121 (NIH)	Screens for 14 thoracic conditions with confidence scores
🔎 Medical Web Search	Ollama web search	Live medical search with clean, sourced summaries
🌐 Full RTL/LTR UI	—	Complete Arabic ↔ English toggle with layout direction switch

---

Architecture

┌─────────────────────────────────────────────────┐
│                  chatbot_ui.html                 │
│           (Single-page bilingual UI)             │
└────────┬──────────┬──────────┬──────────────────┘
         │          │          │          │
    Port 8000  Port 8001  Port 8002  Port 8003
         │          │          │          │
┌────────▼──┐ ┌─────▼──────┐ ┌▼──────────┐ ┌─────▼──────┐
│rag_chatbot│ │Live_MedProc│ │  server.py │ │  main.py   │
│    .py    │ │    .py     │ │ (imaging)  │ │(web search)│
│           │ │            │ │            │ │            │
│ Mistral   │ │ Qwen2.5-VL │ │DenseNet-121│ │   Ollama   │
│ + FAISS   │ │   (HF)     │ │   (NIH)    │ │ web search │
└───────────┘ └────────────┘ └────────────┘ └────────────┘

---

Setup

1. Clone & Configure

git clone https://github.com/Shrouk-Sharaf/Tammeny
cd Tammeny
cp .env.example .env
# Edit .env and add your API keys

2. Install Dependencies

# Root (chatbot + vision)
pip install fastapi uvicorn python-multipart langchain langchain-community \
    langchain-text-splitters langchain-huggingface langchain-mistralai \
    faiss-cpu sentence-transformers pypdf python-dotenv requests

# Imaging service
cd imaging-service && pip install -r requirements.txt && cd ..

3. API Keys

Key	Where to Get	Cost
MISTRAL_API_KEY	console.mistral.ai	Free tier
HUGGINGFACEHUB_API_TOKEN	huggingface.co/settings/tokens	Free
HF_ROUTER_TOKEN	Same HuggingFace token	Free

---

Running

Open 4 terminals:

# Terminal 1 — Chatbot + PDF Q&A
python rag_chatbot.py
# → http://127.0.0.1:8000

# Terminal 2 — Medical Image Analysis
uvicorn Live_MedProc:app --reload --host 127.0.0.1 --port 8001
# → http://127.0.0.1:8001

# Terminal 3 — Chest X-Ray Analysis
cd imaging-service && python server.py
# → http://0.0.0.0:8002

# Terminal 4 (optional) — Web Search
python main.py
# → http://0.0.0.0:8003

Then open chatbot_ui.html in your browser.

---

Docker (Imaging Service)

cd imaging-service
docker build -t tammeny-imaging .
docker run -p 8002:8002 tammeny-imaging

---

API Reference

Chatbot — :8000

Method	Endpoint	Description
GET	/	Serve chatbot HTML UI
POST	/load_pdf/	Upload PDF for RAG context
POST	/chat	Send message {prompt, session_id}

Vision AI — :8001

Method	Endpoint	Description
GET	/	Health check
POST	/analyze_image/	Analyze image (form: image, human_prompt)

X-Ray Service — :8002

Method	Endpoint	Description
GET	/health	Health check
GET	/models	Available models info
POST	/analyze/xray	Analyze X-ray (file, confidence_threshold)

Web Search — :8003

Method	Endpoint	Description
GET	/	Health check
POST	/chat	Search query {prompt}

---

Project Structure

healthcare-ai/
├── rag_chatbot.py          ← Chatbot API (port 8000)
├── Live_MedProc.py         ← Vision AI API (port 8001)
├── main.py                 ← Web-search API (port 8003)
├── chatbot_ui.html         ← Bilingual single-page UI
├── .env.example            ← API key template
├── requirements.txt
│
├── service/imaging-service/
│   ├── server.py           ← Imaging API (port 8002)
│   ├── inference.py        ← Inference pipeline
│   ├── postproc.py         ← Result formatter
│   ├── Dockerfile
│   └── models/xray_model/
│       ├── xray_analyzer.py
│       ├── model_loader.py
│       └── nih_processor.py
│
└── docs/
    └── screenshots/
        ├── chat_eng.png
        ├── xray_high.png
        └── xray_low.png

---

Disclaimer

This is a demonstration project built for educational purposes. It is not intended for clinical use and is not a substitute for professional medical advice, diagnosis, or treatment. Always consult a qualified healthcare professional for medical decisions.

---