Guardrail-Based Radiology Protocolling with Large Language Models: Balancing Automation and Oversight
Journal of Imaging Informatics in Medicine · 12.11.2026
DOI: 10.1007/s10278-026-02496-9 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Imaging Informatics in Medicine, online erschienen am 12.11.2026. Verfasst von Alexander W Marka und 6 weiteren Autor:innen. Beteiligte Einrichtung: Institute for Diagnostic and Interventional Radiology, School of Medicine and Health, TUM Klinikum, Technical University of Munich (TUM), Munich, Ger…. Die Arbeit nennt 21 Literaturangaben.
Abstract (Originalsprache)
Safe clinical deployment of large language models (LLMs) for radiology protocolling requires knowing when not to automate. We evaluated a "guardrail-first" workflow that defers incomplete or safety-critical CT and MRI requests to radiologists before assignment. In this single-center retrospective study, 537 body imaging requests were temporally split into prompt development (n = 109) and validation (n = 428) cohorts. An expert radiologist classified requests as protocol-eligible or requiring review. Six LLM pipelines (Gemini 3.1 Pro, Claude Sonnet 4.6, GPT-5.2, GPT-5.2 high, DeepSeek V3.2, and DeepSeek V3.2 reasoner) were evaluated in two stages (review triage, protocol assignment) against human protocolers. Primary outcome was clinically appropriate assignment (exact/equivalent protocol or correct review escalation), with bootstrapped 95% CIs. The expert reference standard classified 57/428 requests (13.3%) as requiring review (15 safety-critical, 23 major missing information, 19 minor clarification). For the 38 requests with safety-critical issues or major missing information, Gemini 3.1 Pro achieved 100% sensitivity (38/38; 95% CI, 90.7-100), versus 94.7%-97.4% for other pipelines. Among 371 protocol-eligible requests, unnecessary escalation ranged from 1.3% (5/371) to 34.0% (126/371), remaining lowest for reasoning-enabled models. The human benchmark achieved 83.6% (95% CI, 79.8-87.3) exact/equivalent assignment, exceeded by Gemini 3.1 Pro (+12.7 percentage points; 95% CI, 8.6-16.7), Claude Sonnet 4.6 (+12.4; 8.4-16.4), GPT-5.2 high (+11.6; 7.3-15.9), and DeepSeek V3.2 reasoner (+8.6; 4.3-12.9). End-to-end clinically appropriate assignment ranged from 65.0% (60.5-69.4) for GPT-5.2 to 96.3% (94.4-97.9) for Gemini 3.1 Pro; the three highest-performing pipelines did not differ significantly. A guardrail-first LLM protocolling workflow combined high-performance automation with reliable identification of requests requiring radiologist review, supporting further prospective evaluation.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Institute for Diagnostic and Interventional Radiology, School of Medicine and Health, TUM Klinikum, Technical University of Munich (TUM), Munich, Germany.
- Department of Radiology, McGill University Health Centre, Montreal, QC, Canada.
- Department of Medical Imaging, Temerty Faculty of Medicine, University of Toronto, Toronto, ON, Canada.
- Department of Pharmacology, Dalhousie Medicine New Brunswick, Faculty of Medicine, Dalhousie University, Saint John, NB, Canada.
- Department of Translational Medicine, The Hospital for Sick Children, Toronto, ON, Canada.
- Department of Radiology, McGill University Health Centre, Montreal, QC, Canada. dominik.deniffel@mcgill.ca.
Zitieren
APA
Marka, A. W., Reinhold, C., Nassim, D., Pfeiffer, D., Pelage, J. P., Namdar, K., & Deniffel, D. (2026). Guardrail-Based Radiology Protocolling with Large Language Models: Balancing Automation and Oversight. Journal of Imaging Informatics in Medicine. https://doi.org/10.1007/s10278-026-02496-9
Vancouver
Marka AW, Reinhold C, Nassim D, Pfeiffer D, Pelage JP, Namdar K, et al. Guardrail-Based Radiology Protocolling with Large Language Models: Balancing Automation and Oversight. Journal of Imaging Informatics in Medicine. 2026. doi:10.1007/s10278-026-02496-9
BibTeX
@article{Marka2026,
author = {Marka, Alexander W and Reinhold, Caroline and Nassim, David and Pfeiffer, Daniela and Pelage, Jean-Pierre and Namdar, Khashayar and Deniffel, Dominik},
title = {Guardrail-Based Radiology Protocolling with Large Language Models: Balancing Automation and Oversight},
journal = {Journal of Imaging Informatics in Medicine},
year = {2026},
doi = {10.1007/s10278-026-02496-9}
}
EUVIMED