Dynamic TyG trajectories cumulative TyG burden are associated with in-hospital mortality in acute brain injury: a multicenter interpretable machine-learning analysis - PubMed
3 hours ago
- #acute brain injury
- #machine learning
- #metabolic trajectory
- Dynamic TyG trajectories and cumulative TyG burden are associated with in-hospital mortality in acute brain injury (ABI).
- Three TyG trajectories were identified: low-slightly increasing (LSI), moderate-increasing (MI), and persistently high (PH).
- After day 7, both MI and PH trajectories showed higher in-hospital mortality compared to LSI.
- Threshold-based mean area under the curve (TBM) showed a positive association with mortality, remaining significant in fully adjusted models.
- The study utilized a machine-learning model (ExtraTrees) for consistent internal and external validation performance.
- TyG trajectory and TBM8p7 were among the most important predictors alongside SOFA score and vasopressor use.
- The findings suggest TyG is better modeled as a time-aware exposure in ICU-treated ABI.