Explainable AI identifies key biomarkers for acute kidney injury prediction in the ICU
XGBoost identified key predictors of early new-onset AKI and RRT need in the ICU, highlighting both emerging (endostatin, NGAL) and established biomarkers (lactate, albumin), alongside known clinical predictors. It also improved predictive accuracy for both outcomes. Further clinical evaluation of these biomarkers and XAI models is warranted.
Authors
- Hans Friberg (ORCID: https://orcid.org/0000-0002-5588-0098)
- Hazem Koozi (ORCID: https://orcid.org/0000-0002-5892-3524)
- Attila Frigyesi (ORCID: https://orcid.org/0000-0002-0155-4828)
- Jonas Engström (ORCID: https://orcid.org/0000-0003-1421-3640)
Institutions
- Lund University (SE)
- Kristianstad University (SE)
- Skåne University Hospital (SE)
- Centralsjukhuset Kristianstad (SE)
Publication Details
- Journal
- Intensive Care Medicine Experimental
- Published
- 2025-10-22
- DOI
- https://doi.org/10.1186/s40635-025-00816-x
- Citations
- 1
- Primary Topic
- Acute Kidney Injury Research
- Type
- article
- Field-Weighted Citation Impact
- 1.43
Funders
- Skånes universitetssjukhus