LungHeal is a web application that empowers doctors and patients to upload histopathology lung images and receive accurate AI-driven diagnoses using state-of-the-art deep learning models. 🧬✨
📚 We use the publicly available Lung and Colon Cancer Histopathological Images Dataset from Kaggle. This curated dataset enables precise model training across three diagnostic classes:
- Benign
- Malignant Squamous Cell Carcinoma
- Malignant Adenocarcinoma
🤖 LungHeal introduces three AI models for image classification:
- MobileNetV2
- DenseNet121
- InceptionV3
All models are in .keras format for better metadata handling and are available for users to select the most accurate result. This is a major improvement over legacy systems that only used .h format with ResNet.
- 🚀 Added support for multi-model selection with individual prediction outputs.
- 🖼️ Support for histopathology-based diagnosis instead of standard X-rays.
- 👨⚕️ Dual login system for patients and doctors.
- 🌐 Intuitive UI built with React and backend powered by Django.
- User logs in (doctor or patient).
- Upload histopathology lung image.
- Select preferred AI model (MobileNetV2, DenseNet121, or InceptionV3).
- AI returns prediction: Benign, Malignant Squamous, or Malignant Adenocarcinoma.
- Python
- Django (Backend)
- React (Frontend)
- Keras (Model framework)
- HTML/CSS
- TensorFlow
# Clone the repository
git clone https://github.com/jerryblessed/lungheal.git
cd lungheal
# Install backend dependencies
pip install -r requirements.txt
# Run Django server
python manage.py runserver
# In another terminal, run frontend
cd frontend
npm install
npm startVisit http://localhost:3000 to interact with the web app.
Traditional treatment included:
- 💉 Chemotherapy – Uses strong drugs, often with severe side effects.
- 🔪 Surgery – Removes tumors but isn’t always viable.
- ☢️ Radiation – Risky and may affect healthy tissue.
LungHeal provides a non-invasive, accurate, and fast diagnosis option powered by AI.
"It is health that is the real wealth and not pieces of gold and silver."
Made with ❤️ by Team LungHeal – Your AI partner for faster, safer lung cancer detection.