Abstract

The repository (GitHub: https://github.com/yhzhu99/pyehr-playground) is a practical implementation of the arXiv paper: "A Comprehensive Benchmark for COVID-19 Predictive Modeling Using Electronic Health Records in Intensive Care" Introduction The COVID-19 EHR Benchmark is a comprehensive platform for predictive modeling using Electronic Health Records (EHR) in Intensive Care Units (ICUs). This benchmarking effort is the first of its kind for patient-level COVID-19 prediction tasks in ICUs. The main goal is to provide a reliable and standardized benchmark for researchers and practitioners in the field. Features Performance Table: Provides the average score across all folds. View Here Detailed Performance Table: Includes comprehensive performance details for each fold. View Here Checkpoints & Logs: Access all model checkpoints and logs. Access via Google Drive Background To the best of our knowledge, this is the first benchmarking effort for patient-level COVID-19 prediction tasks in ICUs. We have made our code publicly available, enabling others to build complete benchmarks and reproduce all results. Our well-structured data preprocessing and modeling modules can also be easily applied to generate customized tasks and results. The benchmark code and documentations can be accessed at GitHub Repository. Moreover, we have released all the benchmark experiment results and trained models on this online platform, which includes model performances with all hyperparameter combinations for both tasks and makes the results easy to query and download. Further details on data and code availability, including the benchmark code structure, are provided in Supplementary Materials. Access The platform can be accessed at https://pyehr.netlify.app.

Rights

This content is not covered by the Open Government Licence. Please see source record or item for information on rights and permissions.

Cite as

Zhu, Y., Wang, W., Gao, J. & Ma, L. 2024, '[CODE] COVID-19 EHR Benchmark - Online Platform', Zenodo. https://doi.org/10.5281/zenodo.10573646

Downloadable citations

Download HTML citationHTML Download BIB citationBIB Download RIS citationRIS
Last updated: 09 June 2025
Was this page helpful?