This project focuses on the classification of ECG arrhythmias using the MIT-BIH Arrhythmia Database. It includes end-to-end preprocessing of ECG signals, heartbeat labeling according to AAMI standards, feature extraction, and machine learning-based classification. Project Workflow
Data Collection
Load ECG signals from MIT-BIH Arrhythmia Database using wfdb.
? Signal Preprocessing
Remove noise, extract heartbeats, and classify beats using AAMI standards.
Feature Extraction
Extract statistical or signal-based features (RR interval, amplitude, etc.).
? Modeling The models used include:
LSTM (Long Short-Term Memory)
GRU (Gated Recurrent Unit)
Hybrid Systems combining LSTM and GRU
Evaluation
