❤️ ECG-Arrhythmia-Classification-using-Artificial-Neural-Network

📝Noted

This project is following the AMMI Standard which used only 44 out of 48 dataset, the dataset were deleted included 102, 104, 107, 217.

Moreover, the AAMI Standard merged dataset from 15 classes into 5 classes contains Normal beat, Supreventricular Ectopic Beat, Ventricular Ectopic beat, Fusion Beat and Unknown Beat.

📘Training (22 Records)

101, 106, 108, 109, 112, 114, 115, 116, 118, 119, 122, 124, 201, 203, 205, 207, 208, 209, 215, 220, 223, 230

📘Testing (22 Records)

100, 103, 105, 11, 113, 117, 121, 123, 200, 202, 210,212, 213, 214, 219, 221, 222, 228, 231, 232, 233, 234

✨You can using the already split train dataset and testing dataset in hdf5 format. (You don't need to run the preprocessing code and download dataset).

📚Dependency

Numpy

Matplotlib

h5py

Keras 2.4.3

Sklearn 0.22.1

Tensorflow cpu 2.1.0

💾Dataset

The dataset was downloaded from kaggle website; which data store in csv file and annotation file in txt, not original format from physionet website. https://www.kaggle.com/taejoongyoon/mitbit-arrhythmia-database

Train and test results

|

*Test-set result accuracy was over 90%.

💌Achknowledgement

https://www.kaggle.com/taejoongyoon/mitbit-arrhythmia-database

https://www.physionet.org/content/mitdb/1.0.0/