An auditable multimodal agentic system for sequential decision-making across the acute stroke pathway
- 1Minhang Hospital, Fudan University
Abstract
Ischemic stroke is a leading cause of long-term disability worldwide. Although timely reperfusion can improve outcomes, realizing this benefit requires clinicians to integrate fragmented clinical and imaging information with evolving evidence under the time pressure of emergency care, often with limited access to specialist expertise. Here we present StrokeAgent, a multimodal agentic decision-support framework designed around temporal alignment and continuous information synthesis across the acute stroke pathway. It processes clinical records, laboratory results, raw electrocardiographic (ECG) waveforms and source multimodal CT images to support emergency triage, the initial thrombolysis decision and final reperfusion planning. As information becomes available, StrokeAgent formulates and updates recommendations at each decision point, grounding each recommendation in an auditable chain of supporting findings and evidence. We evaluated the system in a curated set of 100 patients with multimodal clinical and imaging data, deliberately enriched for clinically complex reperfusion scenarios alongside standard presentations. StrokeAgent achieved 86% concordance with expert-adjudicated final decisions, exceeding the mean performance of three tool-free large language models (LLMs) and six resident and attending neurologists by 13 and 17 percentage points, respectively, under matched information content and timing. Complete-pathway concordance (agreement at every applicable decision point) was 76%, with even larger margins over both comparator groups. On average, StrokeAgent completed the full decision pathway 73.9 s faster than the case-level median neurologist time. In 1,200 blinded clinician assessments, StrokeAgent received higher overall clinical utility ratings than each LLM (common odds ratios, 10.8-16.3) and higher reasoning-quality ratings, with fewer outputs judged to pose severe potential harm. StrokeAgent demonstrates the feasibility of a decision-support system that integrates native multimodal data processing with auditable, evidence-supported recommendations aligned with the sequential nature of acute stroke care. Prospective studies are needed to determine whether physician-supervised use improves clinical care.
Related
- From Clinical Free Text to Auditable Concepts: An Agentic Framework for Interpretable Prediction
- Human-Guided Agentic AI for End-to-End EHR Research: Sequential Antibody-Drug Conjugate Pathways in Breast Cancer
- Deploying YOLO for whole-slide mitosis detection: A transparent, auditable, and reproducible pipeline for pathology AI-assisted diagnosis and decision-support tool
- Real-world evaluation of a multilingual agentic retrieval-augmented clinical AI framework for evidence-grounded medical decision support
- Understanding digital diffusion and online decision-making in public health promotion: a sequential mixed-methods study of the violence prevention charter in Canada
BibTeX
@misc{yang2026auditable,
title = {{An auditable multimodal agentic system for sequential decision-making across the acute stroke pathway}},
author = {K.-C. Yang and Z. Bao and L.-S. Tong and J. Yuan and Y. Chen and X. Xue and Y.-q. Zhu and Y. Luo and X. Mao and J. Zheng and Z. Liu and Y. Wang and Z. Zhang and R. Ji and X. Shi and J. Peng and Z. Wei and J. Zhao},
howpublished = {medRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.24.26363786},
url = {https://www.medrxiv.org/content/10.64898/2026.09.24.26363786v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - An auditable multimodal agentic system for sequential decision-making across the acute stroke pathway
AU - K.-C. Yang
AU - Z. Bao
AU - L.-S. Tong
AU - J. Yuan
AU - Y. Chen
AU - X. Xue
AU - Y.-q. Zhu
AU - Y. Luo
AU - X. Mao
AU - J. Zheng
AU - Z. Liu
AU - Y. Wang
AU - Z. Zhang
AU - R. Ji
AU - X. Shi
AU - J. Peng
AU - Z. Wei
AU - J. Zhao
PY - 2026
DA - 2026/09/24
PB - medRxiv
DO - 10.64898/2026.09.24.26363786
UR - https://www.medrxiv.org/content/10.64898/2026.09.24.26363786v1
AB - Ischemic stroke is a leading cause of long-term disability worldwide. Although timely reperfusion can improve outcomes, realizing this benefit requires clinicians to integrate fragmented clinical and imaging information with evolving evidence under the time pressure of emergency care, often with limited access to specialist expertise. Here we present StrokeAgent, a multimodal agentic decision-support framework designed around temporal alignment and continuous information synthesis across the acute stroke pathway. It processes clinical records, laboratory results, raw electrocardiographic (ECG) waveforms and source multimodal CT images to support emergency triage, the initial thrombolysis decision and final reperfusion planning. As information becomes available, StrokeAgent formulates and updates recommendations at each decision point, grounding each recommendation in an auditable chain of supporting findings and evidence. We evaluated the system in a curated set of 100 patients with multimodal clinical and imaging data, deliberately enriched for clinically complex reperfusion scenarios alongside standard presentations. StrokeAgent achieved 86% concordance with expert-adjudicated final decisions, exceeding the mean performance of three tool-free large language models (LLMs) and six resident and attending neurologists by 13 and 17 percentage points, respectively, under matched information content and timing. Complete-pathway concordance (agreement at every applicable decision point) was 76%, with even larger margins over both comparator groups. On average, StrokeAgent completed the full decision pathway 73.9 s faster than the case-level median neurologist time. In 1,200 blinded clinician assessments, StrokeAgent received higher overall clinical utility ratings than each LLM (common odds ratios, 10.8-16.3) and higher reasoning-quality ratings, with fewer outputs judged to pose severe potential harm. StrokeAgent demonstrates the feasibility of a decision-support system that integrates native multimodal data processing with auditable, evidence-supported recommendations aligned with the sequential nature of acute stroke care. Prospective studies are needed to determine whether physician-supervised use improves clinical care.
ER -