An interpretable machine learning model for early prediction of subsequent observed CRRT initiation in patients with sepsis-associated acute kidney injury
Yaoyao Tang, Shuai Zhang, Qiujun Wang, Qi Zhang, Liye Ji, Ruoxi Li, Mingxing Fang · 2026
Read An interpretable machine learning model for early prediction of subsequent observed CRRT initiation in patients with sepsis-associated acute kidney injury by Yaoyao Tang, Shuai Zhang, Qiujun Wang, Qi Zhang, Liye Ji, Ruoxi Li, Mingxing Fang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13627203.txt.
Description
Yaoyao Tang and co-authors published this clinical prediction study in 2026. In sepsis-related kidney injury, doctors must decide who will need continuous dialysis, a scarce and intensive treatment. The team built a machine learning model that uses only data from the first twenty-four hours after sepsis diagnosis to predict dialysis starts over the following week. Records came from two large intensive care databases.
Eight algorithms were compared, and a gradient boosting model with seven predictors won. Its inputs included kidney function, urine output, illness severity, and lactate. Accuracy was strong inside the home database and moderate in the external one, though its risk estimates ran high there. The authors stress the model predicts what clinicians did, not who truly needed dialysis. They require prospective testing and local recalibration before any bedside use.
Catalog record
| CREATOR | Yaoyao Tang, Shuai Zhang, Qiujun Wang, Qi Zhang, Liye Ji, Ruoxi Li, Mingxing Fang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |