Patient models distinguish recorded STEMI mortality outcomes
Source check: Source checked. Not independent scientific confirmation.
Finding: A random-forest model evaluated on an internal holdout from1700 STEMI patients reported ROC areas of0.808 for in-hospital mortality and0.783 for30-day mortality. This source-specific evaluation uses clinical and complication-related inputs; it does not establish a pre-treatment warning tool or improved care.
Technology: Random forest; multilayer perceptron; tree ensembles
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Supervised classifiers learn clinical-feature relationships to observed in-hospital and30-day mortality; SHAP describes model-used features.
- Test and result
- Patients were recruited in2021–2023 and later divided1360/340 for training and internal evaluation, stratified by mortality; resampling was limited to training. The cohort had168 in-hospital and210 thirty-day deaths.
- Limitation / still unknown
- Outcome-informed feature screening and data refinement are described before the isolated split, with incomplete nesting detail. Inputs include shock, treatment delay, mechanical complications and echocardiography, so an early pre-treatment prediction time is not established. The paper calls random forest best, but Table3 gives higher30-day ROC areas to other models; no overall-best or significance claim is adopted. Prospective recruitment alone does not establish prospective model validation; no external cohort or treatment benefit was demonstrated.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC12223519/ · https://doi.org/10.1016/j.ihj.2025.03.010
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 21 Nov 2026. Review scope: Actual relevant primary computational/clinical Methods and high-level Results read; no operational biological protocols inspected.
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.