How tharunmellow2007-wq/parkinson-s-disease-screening-system works
This is a Gradio voice-analysis application for Parkinson’s disease detection and four-stage classification. Users record or upload audio, select noise reduction, and receive biomarker results; stage classification then presents clinical descriptions, symptoms, recommendations, and visual references. The runtime is concentrated in app.py, with model artifacts under models/. The README documents preprocessing and analysis capabilities not fully visible in the sampled excerpts, so those components are shown at the documented level; wiring beyond that is omitted.
Main entry points: app.py Disease detection · parkinsons_ensemble_model.pkl Ensemble model · feature_scaler.pkl Feature scaler · feature_info.json Feature metadata
Components 13
User interface
- Gradio interface
app.py - Audio input
app.py
Voice analysis
- Audio preprocessing
app.py - Biomarker extraction
app.py
Classification
- Disease detection
app.py - Stage classification
app.py - Model loading
app.py - Ensemble model
models/parkinsons_ensemble_model.pkl - Feature scaler
models/feature_scaler.pkl - Feature metadata
models/feature_info.json
Clinical results
- Clinical results
app.py - Stage references
images
Other components
- User
Connections 14
- User → Gradio interface: uses
- Gradio interface → Audio input: accepts audio
- Audio input → Audio preprocessing: submits audio
- Audio preprocessing → Biomarker extraction: extracts from
- Biomarker extraction → Disease detection: provides features
- Disease detection → Stage classification: triggers
- Model loading → Ensemble model: loads
- Model loading → Feature scaler: loads
- Model loading → Feature metadata: loads
- Disease detection → Model loading: uses model
- Stage classification → Clinical results: produces results
- Disease detection → Clinical results: produces results
- Clinical results → Stage references: loads references
- Clinical results → Gradio interface: renders