Syarafi/PPG-Arrhythmia

This is a Gradient Boost model that predicts Arrhythmia condition based on PPG signal input in a format of 250Hz.

Format of Input:

PPG_input = []: Array of signal, target is 2500 (10 second PPG signal readings)

Output = { 'success': True/False, 'prediction': prediction, 'arrhythmia_name': arrhythmia_name, 'confidence': confidence, 'quality_score': quality_score (quality of PPG signal), 'all_probabilities': {name: prediction_proba[code] for name, code in ARRHYTHMIA_TYPES.items()} }

There's also a python code for the inference interface called inference.py in the same directory (you'll need to open the model with pickle.load())