A Python-based application that provides a user-friendly interface for Parkinson's Disease detection through spiral drawing analysis, inspired by Robin T. White's research work.
- Simple GUI Interface: Built with Tkinter for easy image upload and analysis
- Real-time Processing: Immediate results after image upload
- Training Data Collection: Optional user feedback system to improve model accuracy
- Image Processing Pipeline: Uses OpenCV and scikit-image for feature extraction
- Machine Learning Model: Random Forest classifier for prediction
PDiagnose/
├── Interface.py # GUI and main application logic
├── parkinson_model.py # ML model and feature extraction
├── process_data.py # Data preprocessing utilities
├── process_images.py # Image processing functions
└── drawings/
└── spiral/
├── training/
└── testing/
- opencv-python
- scikit-image
- scikit-learn
- xgboost
- numpy
- pillow
- imutils
- tkinter
pip install -r requirements.txt- Run the application:
python Interface.py- Draw a spiral pattern according to the guide shown
- Use "Select Image" to upload your drawing
- View analysis results
- Optionally contribute to model training by confirming diagnosis status
The GUI application provides:
- Visual guide for spiral drawing
- Real-time image analysis
- Prediction results display
- Optional feedback collection
- Training data management
This is a proof-of-concept implementation based on research by Robin T. White. Not for medical diagnosis.
MIT License