Triage-Time Machine Learning for Computed Tomography Acquisition Prioritization in the Emergency Department
Academic Radiology · 2026-11 · 33 (11) · 5122-5131
DOI: 10.1016/j.acra.2026.08.017 ↗Kurz gefasst
Fachartikel in der Zeitschrift Academic Radiology, 2026-11. Verfasst von Philip Jarrett und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Emergency Medicine, UT Southwestern Medical Center, Dallas, TX, United States (P.J., D.D.). Electronic address: philip.jarrett@utsouthw…. Hauptthemen laut MEDLINE: Triage, Tomography, X-Ray Computed, Emergency Service, Hospital, Machine Learning. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 20 Literaturangaben. Gefördert von: The University of Texas Southwestern Medical Center.
Abstract (Originalsprache)
RATIONALE AND OBJECTIVES: To evaluate whether triage-time machine learning (ML) can improve computed tomography (CT) acquisition prioritization beyond first-in-first-out (FIFO) ordering and the Emergency Severity Index (ESI). MATERIALS AND METHODS: Using two emergency department datasets (Site A: 118,385 visits; Site B: 425,087 visits), we identified 43,051 and 57,728 CT patients. Gradient-boosted classifiers trained on 32 triage-time features predicted 6-hour deterioration. Because true concurrent queues could not be reconstructed, we simulated 10-, 15-, and 20-patient CT queues. Shifts were sampled until 1000 deteriorator-positive shifts were accrued per queue-size scenario. The primary endpoint was the mean deteriorator queue percentile; the secondary endpoint was the first-quartile recall. A fully crossed experiment assessed transferability. RESULTS: Under FIFO, the mean deteriorator queue percentile remained near mid-queue at both sites (44.9%-47.4%). Site-specific ML reduced this to 14.2%-15.2% at Site A and 15.7%-17.1% at Site B, versus 21.2%-22.6% and 32.9%-34.8% for ESI. First-quartile recall with ML was 75.0%-78.4% at Site A, and 72.1%-76.1% at Site B. Cross-site transfer remained weakest for the Site A model applied to Site B (29.7%-30.9%). Internal models were well calibrated. At Site A, 72.8% of deteriorating CT patients had at least one affirmed actionable CT finding. CONCLUSION: In simulation, triage-time ML robustly improves CT scanner access across plausible congestion levels. ESI captures part of the available signal, but local retraining is preferred because cross-site transfer degrades performance.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Emergency Medicine, UT Southwestern Medical Center, Dallas, TX, United States (P.J., D.D.). Electronic address: philip.jarrett@utsouthwestern.edu.
- Department of Radiology, UT Southwestern Medical Center, Dallas, TX, United States (P.E.d.A.K.).
- Department of Emergency Medicine, UT Southwestern Medical Center, Dallas, TX, United States (P.J., D.D.).
- Lyda Hill Department of Bioinformatics, UT Southwestern Medical Center, Dallas, TX, United States (M.H., A.J.).
Förderung
- The University of Texas Southwestern Medical Center
Zitieren
APA
Jarrett, P., Kuriki, P. E. A., Diercks, D., Holcomb, M., & Jamieson, A. (2026). Triage-Time Machine Learning for Computed Tomography Acquisition Prioritization in the Emergency Department. Academic Radiology, 33(11), 5122-5131. https://doi.org/10.1016/j.acra.2026.08.017
Vancouver
Jarrett P, Kuriki PEA, Diercks D, Holcomb M, Jamieson A. Triage-Time Machine Learning for Computed Tomography Acquisition Prioritization in the Emergency Department. Academic Radiology. 2026;33(11):5122-5131. doi:10.1016/j.acra.2026.08.017
BibTeX
@article{Jarrett2026,
author = {Jarrett, Philip and Kuriki, Paolo Eduardo de Aguiar and Diercks, Deborah and Holcomb, Michael and Jamieson, Andrew},
title = {Triage-Time Machine Learning for Computed Tomography Acquisition Prioritization in the Emergency Department},
journal = {Academic Radiology},
year = {2026},
volume = {33},
number = {11},
pages = {5122--5131},
doi = {10.1016/j.acra.2026.08.017}
}
EUVIMED