ECG Classification

ECG Classification based on MLP RNN LSTM Attention-Model CNN

Materials

• MIT Arrythmia database and MIT Normal Sunis Database
• CCDD database http://www.ecgdb.com

Introduce

• The DeHaze folder is a dehaze model of image
• EEG folder is a EEG classification model
• other ECG model folder contains some simple models or some ideas for trying
• 12-Lead ECG model is four deep learning model which build with pytorch
  • Vanilla-CNN is a simple CNN model to classify the CCDD database
  • Channel-RNN is a CNN+RNN network
  • Featrue-CNN is a RNN+CNN network
  • Multi-RNN is a 12-Lead based RNN network

Conclusion

ECG signals were classified using different deep learning models. And try to combine LSTM with CNN to process multi-lead sequence signals. The model performance is not particularly good, but I hope these idea will help you a little.