Bsingstad/ECG-age

Public repository associated with: "Using deep convolutional neural networks to predict patients age based on ECGs from an independent test cohort"

Age estimation using AI on 12-lead ECGs

Reproduce experiments

To reproduce the experiments in this GitHub repository, you should run the Jupyter Notebooks in the following order:

• cross_validation.ipynb
• final_training.ipynb
• testset_performance.ipynb

Run Jupyter Notebooks in Google Colab

To run the Jupyter Notebooks you will need GPUs. For those of you who don't have your own GPU, we thought that it would be convenient to use free GPUs from Google.

The easiest way to open the Jupyter Notebooks from this GitHub repository in Google Colab is to:

1. Use a Google Chrome browser
2. Install the Google Colab extention: https://chrome.google.com/webstore/detail/open-in-colab/iogfkhleblhcpcekbiedikdehleodpjo
3. Click on the Notebook, in this GitHub repository, you want to open
4. Finally click on your new Google Colab extention right underneath the tabs in your Google Chrome browser.

Paper

Using deep convolutional neural networks to predict patients age based on ECGs from an independent test cohort

Citation

Please acknowledge our work by citing our conference paper:

@inproceedings{singstad_2022,
  title={Using deep convolutional neural networks to predict patients age based on ECGs from an independent test cohort},
  author={Singstad, Bjorn-Jostein and Tawashi, Bilal},
  journal={Proceedings of the Northern Lights Deep Learning Workshop},
  pages={2022--10},
  year={2022},
  publisher={Proceedings of the Northern Lights Deep Learning Workshop}
}