Multi_disease-chest-xray-ai

🫁 Multi-Disease Chest X-ray AI System

A modular deep learning system that runs multiple independent chest X-ray models — Tuberculosis, COVID-19, and NIH Multi-Disease — in parallel on a single input image, with PDF report generation and Grad-CAM explainability.

⚠️ Research / Educational Use Only. Not intended for clinical diagnosis.

---

Features

• Three independent disease-specific models running in parallel with no label interference
• Grad-CAM explainability for TB and COVID-19 detection (heatmap overlays)
• Multi-label prediction for 14 NIH chest pathologies using sigmoid output
• Patient information capture — ID, name, age, gender, referring physician
• Multi-image support — analyze up to 3 X-rays per session
• Automated PDF report generation with findings, heatmaps, and summary table
• Streamlit-based UI with a clean black-and-white clinical aesthetic
• Fully local deployment — no cloud dependency, VS Code ready

---

System Overview

The application orchestrates three independent inference pipelines:

Each model is trained, loaded, and executed independently to preserve interpretability and avoid cross-label interference.

---

Project Structure

├── app.py                  # Main Streamlit application
├── tb/
│   ├── infer.py            # TB inference pipeline
│   └── grad_cam.py         # Grad-CAM for TB model
├── covid/
│   └── infer.py            # COVID-19 inference pipeline
├── multi_disease/
│   └── infer.py            # NIH multi-disease inference pipeline
├── requirements.txt
└── README.md

---

How to Run

1. Clone the repository

git clone https://huggingface.co/devikand/Multi_disease-chest-xray-ai
cd Multi_disease-chest-xray-ai

2. Install dependencies

pip install -r requirements.txt

3. Launch the app

streamlit run app.py

Then open your browser at http://localhost:8501

---

Usage

1. Fill in the Patient Information fields (Patient ID is required)
2. Upload 1–3 chest X-ray images (PNG, JPG, or JPEG)
3. Click "Analyze All X-rays"
4. View per-image results for TB, COVID-19, and NIH pathologies
5. Download the auto-generated combined PDF report

---

Output

For each X-ray, the system produces:

• TB classification result with confidence score and Grad-CAM heatmap
• COVID-19 classification result with confidence score and heatmap (where applicable)
• NIH multi-disease probabilities with a bar chart visualization
• A downloadable PDF medical report including all findings, images, and a summary table

---

Dependencies

Key libraries used:

• streamlit — UI framework
• torch / torchvision — model inference
• opencv-python — image processing
• matplotlib — visualizations
• reportlab — PDF report generation

Install all via:

pip install -r requirements.txt

---

Models

The three models are loaded independently at inference time. Model weights should be placed in their respective module directories (tb/, covid/, multi_disease/) or configured within each infer.py file.

---

Limitations

• This system is built for research and educational purposes only
• Model performance depends on training data distribution — results may not generalize to all clinical settings
• Grad-CAM heatmaps are approximations and should not be used as sole diagnostic indicators
• Not validated for clinical deployment

---

License

This project is intended for academic and research use. Please review model licenses before any commercial application.