Skip to content

Latest commit

 

History

History
120 lines (95 loc) · 4.15 KB

README.md

File metadata and controls

120 lines (95 loc) · 4.15 KB

Organ Classification AI Model

This project focuses on creating a deep learning-based AI model to classify medical images of organs into one of four categories: Eye, Breast, Brain, and Limbs. The model uses TensorFlow and is paired with a user-friendly PyQt5-based GUI to allow seamless interaction.

Features

  • Model Architecture: A custom CNN built using TensorFlow to classify images into four organ categories.
  • User Interface: A dynamic GUI created with PyQt5 to:
    • Upload an image for classification.
    • Display the predicted organ class with confidence levels.
    • Provide notifications for incorrect inputs.
  • Training Pipeline: Organized code structure for training the model on your dataset.
  • Dataset Management: Structured dataset folder for training, validation, and testing.

Project Structure

organ-classifier/
├── app/
│   ├── src/
│   │   ├── main.py                  # Main entry point for the app, runs the PyQt5 GUI
│   │   ├── model.py                 # Contains the model architecture and loading code
│   │   ├── dataset_utils.py         # Functions related to loading or processing the dataset
│   │   ├── predict.py               # Functionality to handle predictions and model inference
│   ├── assets/                      # Images, icons, and other UI assets
│   │   └── logo.png                 # Example logo for the UI (optional)
│   ├── saved_models/                # Folder to store saved models (e.g., trained model weights)
│   │   └── trained_model.h5         # Trained TensorFlow model file
│   ├── requirements.txt             # List of dependencies needed to run the project
│   └── README.md                    # Project documentation (this file)
└── dataset/                         # Dataset folder containing images for training and testing
    ├── train/                       # Training data
    │   ├── Eye/
    │   ├── Breast/
    │   ├── Brain/
    │   └── Limbs/
    ├── validate/                    # Validation data
    │   ├── Eye/
    │   ├── Breast/
    │   ├── Brain/
    │   └── Limbs/
    └── test/                        # Test data
        ├── Eye/
        ├── Breast/
        ├── Brain/
        └── Limbs/

Installation

  1. Clone the repository:

    git clone https://github.com/yourusername/organ-classifier.git
cd organ-classifier

  2. Install the required dependencies:

    pip install -r app/requirements.txt

  3. Ensure you have Python 3.8 or later installed.

Usage

Training the Model

  1. Place your dataset into the dataset/ folder, ensuring it follows the structure outlined above.
  2. Modify any hyperparameters or paths in train.py as needed.
  3. Run the training script: 
    python app/src/train.py

Running the Application

  1. Ensure the trained model (trained_model.h5) is in the saved_models/ directory.
  2. Launch the GUI: 
    python app/src/main.py
  3. Use the GUI to upload images and view predictions.

Dataset Preparation

  • The dataset should be organized into train, validate, and test directories.
  • Each category (Eye, Breast, Brain, Limbs) should have its own subdirectory containing images.
  • Example: 
    dataset/
├── train/
│   ├── Eye/
│   │   ├── image1.jpg
│   │   └── image2.jpg
│   ├── Breast/
│   └── ...
└── test/
    ├── Eye/
    └── ...

Notes

  • Ensure all images are preprocessed to the required input size for the model (default is 224x224).
  • The model's output includes the predicted class and the associated confidence score.

Future Improvements

  • Incorporate additional organ classes.
  • Enhance the GUI for better user experience.
  • Add real-time image preprocessing and augmentation.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Feel free to contribute to the project by submitting pull requests or reporting issues!