alexandrazajdel / wiet_ecg_classification_app

ECG Analysis APP

Final project created as a part of Medical Informatics course at AGH UST in 2020/21 academic year.

Features

Application features:

• ECG signals interactive visualizations,
• ECG heartbeat signals classification using Conv1D neural network,
• heart rate calculating.

Setup

Project is written in Python 3.

1. Install requirements.
2. Add training and test data into data folder.
   Data source: https://www.kaggle.com/shayanfazeli/heartbeat
   Files: mitbih_train.csv, mitbih_test.csv, ptbdb_abnormal.csv, ptbdb_normal.csv
3. Go to src and run app:

streamlit run app.py

Trained models are saved in model folder.

Authors

Maciej Banaś
Aleksandra Pestka