AT-CXR: Uncertainty-Aware Agentic Triage for Chest X-rays
Preprint
Abstract
Agentic AI is advancing rapidly, yet truly autonomous medical-imaging triage, where a system decides when to stop, escalate, or defer under real constraints, remains relatively underexplored. To address this gap, we introduce AT-CXR, an uncertainty-aware agent for chest X-rays. The system estimates per-case confidence and distributional fit, then follows a stepwise policy to issue an automated decision or abstain with a suggested label for human intervention. We evaluate two router designs that share the same inputs and actions: a deterministic rule-based router and an LLM-decided router. Across five-fold evaluation on a balanced subset of NIH ChestX-ray14 dataset, both variants outperform strong zero-shot vision-language models and state-of-the-art supervised classifiers, achieving higher full-coverage accuracy and superior selective-prediction performance, evidenced by a lower area under the risk-coverage curve (AURC) and a lower error rate at high coverage, while operating with lower latency that meets practical clinical constraints. The two routers provide complementary operating points, enabling deployments to prioritize maximal throughput or maximal accuracy. Our code is available at https://github.com/XLIAaron/uncertainty-aware-cxr-agent.Keywords
All Related Versions
This publication has 8 references indexed in Scilit:
- Patho-AgenticRAG: Towards Multimodal Agentic Retrieval-Augmented Generation for Pathology VLMs via Reinforcement LearningProceedings of the AAAI Conference on Artificial Intelligence, 2026
- Agentic AI Framework for End-to-End Medical Data Inference2025
- Agentic AI: Autonomous Intelligence for Complex Goals—A Comprehensive SurveyIEEE Access, 2025
- Real-World evaluation of an AI triaging system for chest X-rays: A prospective clinical studyEuropean Journal of Radiology, 2024
- MDAgents: An Adaptive Collaboration of LLMs for Medical Decision-Making2024
- MMedAgent: Learning to Use Medical Tools with Multi-modal Agent2024
- LLaVA-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day2023
- Computational Radiomics System to Decode the Radiographic PhenotypeCancer Research, 2017