Researchrandom forestcoagulationartificial livermimic database
Artificial Liver Support Improves Coagulation, Predicts Risk
8.0
Researchers conducted a systematic review/meta-analysis of 18 studies (n=1,771) and trained machine-learning models on MIMIC ICU data, published in 2025. The meta-analysis showed ALSS significantly improved INR, PT, APTT, and fibrinogen (all P<.05) with modality-specific effects; an ICU-derived random forest model achieved AUC 92.12%, using dynamic INR to predict coagulation dysfunction for early clinical alerts.
Key Points
- 1Showed ALSS significantly improves INR, PT, APTT, and fibrinogen across 18 studies (n=1,771).
- 2Demonstrated modality-specific efficacy, indicating PE, MARS and other ALSS types differ in coagulation impact.
- 3Delivered ICU-derived random forest model (AUC 92.12%) using dynamic INR to enable early coagulation risk alerts.
Scoring Rationale
Combines meta-analysis and a high-performing ICU-derived ML model, offering actionable clinical prediction but limited broad novelty beyond hepatology.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
