arXiv AI

Examining the Vulnerability of Multi-Agent Medical Systems to Human Interventions for Clinical Reasoning

The paper investigates how human interventions at specific fault points—moments when an AI agent’s reasoning is most vulnerable—affect the diagnostic accuracy of multi‑agent medical systems. Using the MedQA dataset, the authors found that correct interventions can boost baseline accuracy by up to 40%, whereas incorrect or bias‑related interventions can reduce performance by up to 6% and increase diagnostic drift and uncertainty. The study also highlights behavioral parallels between cognitive biases observed in simulated agent conversations and real‑world clinical practice, such as premature closure and susceptibility to misleading cues.

arXiv AI
Jul 1

Agentic AI Enhances Physician Trust in Clinical Decision Making

arXiv:2606. 30658v1 Announce Type: cross Abstract: Medical AI has shifted from reasoning to agentic AI, a new paradigm that autonomously invokes external tools during reasoning, rendering intermediate reasoning steps and tool outputs transparent to users.

By Zhiling Yan, Zhe Fang, David J King, Ann Pongsakul, Eashan Adhikarla, Hui Ren, Sunyang Fu, Quanzheng Li, Lifang He, Xiang Li, Hongfang Liu, Yonghui Wu, Lichao Sun
arXiv AI
Jul 29

PatientAgentBench: A Benchmark Framework for Evaluating Patient-Facing Health AI Agents

arXiv:2607. 25485v1 Announce Type: new Abstract: Health AI is evolving from answering questions to agentic systems that converse with patients, reason about health records, and act on their behalf.

By Korosh Vatanparvar, Ashutosh Joshi, Maria Xenochristou, Mohammad Abuzar Hashemi, Prasad Kasu, Deepak Bansal, Daniel Lopez-Martinez, Anchal Nema, Ramya Ganesan, Will Kimbrough, Alex Woody, Yadunandana Rao, Dilek Hakkani-Tur, Wilko Schulz-Mahlendorf
arXiv AI
Jul 15

First, do NOHARM: a medical safety benchmark and randomized study of physician and AI teaming on clinical consultations

arXiv:2512. 01241v4 Announce Type: replace-cross Abstract: Large language models (LLMs) and medical AI tools are routinely used by physicians and patients for medical advice, yet their clinical safety profiles remain poorly characterized.

By David Wu, Fateme Nateghi Haredasht, Saloni Kumar Maharaj, Priyank Jain, Jessica Tran, Matthew Gwiazdon, Arjun Rustagi, Jenelle Jindal, Jacob M. Koshy, Vinay Kadiyala, Anup Agarwal, Bassman Tappuni, Brianna French, Sirus Jesudasen, Christopher V. Cosgriff, Rebanta Chakraborty, Jillian Caldwell, Susan Ziolkowski, David J. Iberri, Robert Diep, Rahul S. Dalal, Kira L. Newman, Kristin Galetta, J. Carl Pallais, Nancy Wei, Kathleen M. Buchheit, David I. Hong, Vartan Pahalyants, Ernest Y. Lee, Allen Shih, Tamara B. Kaplan, Vishnu Ravi, Sarita Khemani, Thomas A. Buckley, April S. Liang, Daniel Shirvani, Advait Patil, Nicholas Marshall, Kanav Chopra, Joel Koh, Adi Badhwar, Anastasia Perez, Austin J. Schoeffler, Mahbuba Tusty, Chase M. Walton, Liam G. McCoy, David J. H. Wu, Yingjie Weng, Sumant Ranji, Kevin Schulman, Nigam H. Shah, Jason Hom, Arnold Milstein, Arjun K. Manrai, Adam Rodman, Jonathan H. Chen, Ethan Goh
arXiv Computation and Language
Oct 19

The Model Knows Better: Defending Medical Multi-Agent Systems Against Prompt Injection

The paper investigates prompt injection attacks on medical multi‑agent systems that use large language models (LLMs) for clinical decision support. Across two systems, two model families, and three benchmarks, a single injected passage can cause incorrect decisions in up to 61% of cases, and existing defenses only partially mitigate the threat. The authors propose a training‑free defense that verifies each clinical judgment against the patient’s findings, which reduces attack success on MedQA to 4‑7% while preserving accuracy on non‑attacked cases.

By Fardin Ahsan Sakib, Ziwei Zhu, Ozlem Uzuner
arXiv AI
Aug 20

Adaptive Memory and Reflection Multi-Agent System for Medical Question Answering

The paper introduces an adaptive memory and reflection (AMR) multi‑agent system for medical question answering. Each agent has dedicated memory and uses reflection‑based feedback to retrieve relevant prior cases, improving reasoning. The system routes questions through solo, collaborative, or escalated workflows and includes consensus and ethical overseer modules, achieving strong performance on MedQA and MedMCQA datasets.

By Pradeep Murugesan, Luoxiao Yang, Xueli Chen, Xinqi Fan
arXiv AI
Aug 25

MACD: Multi-Agent Clinical Diagnosis with Self-Learned Knowledge for LLM

The paper introduces MACD, a Multi-Agent Clinical Diagnosis framework that enables large language models to self‑learn clinical knowledge through a multi‑agent pipeline of summarization, refinement, and application. MACD is extended into a human‑AI collaborative workflow where multiple diagnostician agents consult iteratively, guided by a judge agent and human oversight. Evaluation on the MIMIC‑MACD cohort shows significant gains in diagnostic accuracy—an average 11.6 percentage‑point improvement over authoritative knowledge for open‑weight LLMs and an 18.3‑percentage‑point boost over physician‑only diagnosis in text‑only vignettes.

By Wenliang Li, Rui Yan, Xu Zhang, Li Chen, Hongji Zhu, Jing Zhao, Junjun Li, Mengru Li, Wei Cao, Zihang Jiang, Wei Wei, Kun Zhang, Shaohua Kevin Zhou