Carotid Doppler Audio Dataset for Stroke Risk Prediction: 149 Recordings from 75 Patients

Doppler audio analysis framework extracting 24 hemodynamic features with Smart Score (1-5) and Risk Score (0-1). Open dataset: 149 recordings from 75 patients with clinical metadata for stroke screening.
Doppler_v1/
├── README.md # Project overview, setup, usage, citation
├── LICENSE # MIT License
├── .gitignore # Excludes large dataset files
├── requirements.txt # Python dependencies
├── app.py # Main Flask web application
├── app.ipynb # Jupyter Notebook (algorithm development)
├── templates/
│ └── index.html # Web interface (Arabic/English)
├── PROCESSED/
│ ├── audio/ # Dataset: 149 WAV recordings (external link)
│ ├── spectrum_images/ # Diagnostic PNG figures
│ └── Result_Export/ # CSV/Excel reports with 24 features
├── dataset/
│ └── metadata.csv # Clinical annotations per recording
└── docs/
└── user_manual.pdf # Platform usage guide