Machine learning predicts mortality in septic patients using only routinely available ABG variables: a multi-centre evaluation.
Source check: Source checked. Not independent scientific confirmation.
Finding: An LSTM using arterial blood gas measurements from the first 48 ICU hours predicted 96-hour mortality among septic ICU stays, with AUC 0.88 in multicenter eICU and 0.85 in single-center MIMIC-III, compared with logistic regression and SOFA. This supports a bounded retrospective prognostic result.
Technology: White Rose accepted version Methods/Results read: retrospectively selected eICU (multi-center) and MIMIC-III cohorts, ABG inputs, LSTM and baseline comparisons; no prospective outcome/clinical benefit claim. No same-result catalog match.
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- White Rose accepted version Methods/Results read: retrospectively selected eICU (multi-center) and MIMIC-III cohorts, ABG inputs, LSTM and baseline comparisons; no prospective outcome/clinical benefit claim. No same-result catalog match.
- Test and result
- White Rose accepted version Methods/Results read: retrospectively selected eICU (multi-center) and MIMIC-III cohorts, ABG inputs, LSTM and baseline comparisons; no prospective outcome/clinical benefit claim. No same-result catalog match.
- Limitation / still unknown
- White Rose accepted version Methods/Results read: retrospectively selected eICU (multi-center) and MIMIC-III cohorts, ABG inputs, LSTM and baseline comparisons; no prospective outcome/clinical benefit claim. No same-result catalog match.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
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: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..
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 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.