Reproduce “Electrocardiogram-based diagnosis of liver diseases: an externally validated and explainable machine learning approach” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Juan Miguel Lopez Alcaraz, Wilhelm Haverkamp, Nils Strodthoff
This work demonstrates that machine learning models leveraging standard ECG features and demographic data can accurately support the diagnosis of various liver diseases. The models exhibit strong predictive performance and generalizability across diverse patient cohorts, while also providing explainable insights into key contributing factors like the QTc interval.
Key insights
- Standard ECG features, particularly the QTc interval, are significant predictors of liver diseases and serve as potential non-invasive biomarkers, reflecting known cardiovascular-hepatic interactions.
- Combining ECG features with basic demographic data (age, sex) yields statistically significant improvements in predictive power compared to using either modality alone for liver disease detection.
- The explainability analysis reveals specific physiological patterns in ECG changes (e.g., QTc prolongation, shifts in RR interval and T-wave axis) linked to different liver conditions, enhancing mechanistic understanding.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.