Skip to content

Latest commit

 

History

History
41 lines (32 loc) · 2.11 KB

README.md

File metadata and controls

41 lines (32 loc) · 2.11 KB

LCCNN

CNN models for MIT-BIH Arrhythmia Database and Open-Source-Event-Driven-ECG-Dataset by PyTorch.

How to train

  1. Download MIT-BIH Arrhythmia Database and Open-Source-Event-Driven-ECG-Dataset.
  2. pip install -r requirements.txt
  3. Rewrite setting_path.py.
  4. Run preprocessing.py
  5. Rewrite and Run 'train.py'

Models

  • CNN
    • Simple CNN model for MIT-BIH Arrhythmia Database designed in [4].
  • LCCNN
    • Simple CNN model for Open-Source-Event-Driven-ECG-Dataset designed in [4].
  • LCCNNLight
    • CNN model for Open-Source-Event-Driven-ECG-Dataset lighter than LCCNN by using additional CNN layer.
  • Resnet34
    • 1d-ResNet34 for MIT-BITH Arrhythmia Database
    • You can use this for Open-Source-Event-Driven-ECG-Dataset by small change.

Input Layers

  • NormalCNN
    • Simple 2ch-CNN layer for Open-Source-Event-Driven-ECG-Dataset designed in [4].
  • TimeEmbedding
    • Input layer for Open-Source-Event-Driven-ECG-Dataset that is 1ch-CNN layer + full-learnable positional encoding layer.
  • TimeSin
    • Input layer for Open-Source-Event-Driven-ECG-Dataset that is 1ch-CNN layer + sinusoidal positional encoding layer.
  • NonTimed
    • 1ch-CNN layer for Open-Source-Event-Driven-ECG-Dataset.

References

[1] https://physionet.org/content/mitdb/1.0.0/
[2] Moody GB, Mark RG. The impact of the MIT-BIH Arrhythmia Database. IEEE Eng in Med and Biol 20(3):45-50 (May-June 2001). (PMID: 11446209)
[3] Goldberger, A., Amaral, L., Glass, L., Hausdorff, J., Ivanov, P. C., Mark, R., ... & Stanley, H. E. (2000). PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals. Circulation [Online]. 101 (23), pp. e215–e220.
[4] M. Saeed et al., "Evaluation of Level-Crossing ADCs for Event-Driven ECG Classification," in IEEE Transactions on Biomedical Circuits and Systems, vol. 15, no. 6, pp. 1129-1139, Dec. 2021, doi: 10.1109/TBCAS.2021.3136206.