Proposing an Approach for Automatic Detection of Atrial Arrhythmia Using Deep Learning Neural Networks

This project consists of three parts: preprocessing(denoising), feature extraction , classification.

The preproccesing step consists of wavelet & median filter for baseline wander removal and also high frequency noises. Then for feature extraction morphological features were extracted and the last part includes deep learning nueral network in order to automaticly  detect atrial arrhythmia in ECG signal.