hocine-drioueche / ECG-classification

ECG-classification

(https://github.com/askeur/ECG-classification)

This Streamlit application is designed to classify heart rhythms in ECG (Electrocardiogram) signals. It provides the following features:

• Uploading and preprocessing ECG files in the .atr, .hea, and .dat formats.
• Classification of ECG signals using different machine learning models.
• Visualizations of ECG signal segments, class distributions, and classification results.

Getting Started

Prerequisites

Before running the application, make sure you have the required Python packages installed. You can install them using pip:

• pip install streamlit
• pip install pandas
• pip install seaborn
• pip install wfdb
• pip install numpy
• pip install scikit-learn
• pip install tensorflow
• pip install matplotlib

Running the Application

To run the ECG Classification Application, open your terminal and navigate to the application's directory. Then, run the following command:

streamlit run app/app.py

This will start the Streamlit development server, and you can access the application in your web browser.

Usage

1. Upload ECG files in the .atr, .hea, and .dat formats -- > csv_train, csv_test
2. Choose a contect ( Training MIH, Training PTB) and a classification model (WaveNet, SVM, KNN, LogisticRegression).
3. Perform preprocessing on the uploaded files.
4. Visualize ECG signals, class distributions, and descriptive statistics.
5. Perform ECG signal classification.
6. View classification reports, confusion matrices, and error analysis.

Model Information

The models used in this application are provided in zip format due to their large size. You need to unzip them before using them.

Built With

• Python
• Streamlit
• scikit-learn
• TensorFlow
• Pandas
• WFDB (WaveForm Database)

Authors

Nabila ASKEUR , KUETE DONGMO Miryam , Abdelkader DEBBAGHI , Hocine DRIOUECHE

Acknowledgments

• Icons and gif (Adob photoshop)

Additional Information

For more information, please refer to the GitHub repository.

Data

This dataset combines two collections of cardiac signals: the MIT-BIH Arrhythmia Dataset and The PTB Diagnostic ECG Database. The number of samples from these two collections is significant, allowing for training deep neural networks.

These signals have undergone certain transformations and have been segmented for study. Each signal segment corresponds to a heart beat. Link to Dataset

Benchmark

Preprocessing, Deep Learning Architectures, Transfer Learning

References

Simplified Understanding of Artificial Neural Networks with 1-D ECG Biomedical Data ECG Research by DataSci You can add this information to your README to provide details about the dataset and references used in your project.