Automated Cardiovascular Record Retrieval by Multimodal Learning between Electrocardiogram and Clinical Report
Run the following Python script to convert ECG signals into images
python ecg_extract_lead_image.py --style=grid --method=gaf,rp,mtf --output_path=data/output/abnormalPTB-XL_Grid
You can modify the visualization style by using the --style argument. The valid options for this argument are concat and grid.
Moreover, you can customize the image visualization approach by adjusting the --methods argument. Make sure to separate multiple methods with a comma, for example gaf,rp,mtf
If you feel our code or models help in your research, kindly cite our papers:
@inproceedings{Qiu2023AutomatedCR,
title={Automated Cardiovascular Record Retrieval by Multimodal Learning between Electrocardiogram and Clinical Report},
author={Jielin Qiu and Jiacheng Zhu and Shiqi Liu and William Jongwon Han and Jingqi Zhang and Chaojing Duan and Michael Rosenberg and Emerson Liu and Douglas Weber and Ding Zhao},
journal={Proceedings of the 3nd Machine Learning for Health symposium, PMLR},
year={2023},
}
@article{qiu2023converting,
title={Converting ECG Signals to Images for Efficient Image-text Retrieval via Encoding},
author={Qiu, Jielin and Zhu, Jiacheng and Liu, Shiqi and Han, William and Zhang, Jingqi and Duan, Chaojing and Rosenberg, Michael and Liu, Emerson and Weber, Douglas and Zhao, Ding},
journal={arXiv preprint arXiv:2304.06286},
year={2023}
}