Toward the Autonomous AI Doctor: Quantitative Benchmarking of an Autonomous Agentic AI Versus Board-Certified Clinicians in a Real World Setting

• URL: http://arxiv.org/abs/2507.22902v1
• Date: Fri, 27 Jun 2025 19:04:44 GMT
• Title: Toward the Autonomous AI Doctor: Quantitative Benchmarking of an Autonomous Agentic AI Versus Board-Certified Clinicians in a Real World Setting
• Authors: Hashim Hayat, Maksim Kudrautsau, Evgeniy Makarov, Vlad Melnichenko, Tim Tsykunou, Piotr Varaksin, Matt Pavelle, Adam Z. Oskowitz,
• Abstract summary: Globally we face a projected shortage of 11 million healthcare practitioners by 2030.<n>No end-to-end autonomous large language model (LLM)-based AI system has been rigorously evaluated in real-world clinical practice.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Background: Globally we face a projected shortage of 11 million healthcare practitioners by 2030, and administrative burden consumes 50% of clinical time. Artificial intelligence (AI) has the potential to help alleviate these problems. However, no end-to-end autonomous large language model (LLM)-based AI system has been rigorously evaluated in real-world clinical practice. In this study, we evaluated whether a multi-agent LLM-based AI framework can function autonomously as an AI doctor in a virtual urgent care setting. Methods: We retrospectively compared the performance of the multi-agent AI system Doctronic and board-certified clinicians across 500 consecutive urgent-care telehealth encounters. The primary end points: diagnostic concordance, treatment plan consistency, and safety metrics, were assessed by blinded LLM-based adjudication and expert human review. Results: The top diagnosis of Doctronic and clinician matched in 81% of cases, and the treatment plan aligned in 99.2% of cases. No clinical hallucinations occurred (e.g., diagnosis or treatment not supported by clinical findings). In an expert review of discordant cases, AI performance was superior in 36.1%, and human performance was superior in 9.3%; the diagnoses were equivalent in the remaining cases. Conclusions: In this first large-scale validation of an autonomous AI doctor, we demonstrated strong diagnostic and treatment plan concordance with human clinicians, with AI performance matching and in some cases exceeding that of practicing clinicians. These findings indicate that multi-agent AI systems achieve comparable clinical decision-making to human providers and offer a potential solution to healthcare workforce shortages.

Related papers

• AI-based Clinical Decision Support for Primary Care: A Real-World Study [1.2764851761863103]
  We evaluate the impact of large language model-based clinical decision support in live care.<n>We studied AI Consult, a tool that serves as a safety net for clinicians by identifying potential documentation and clinical decision-making errors.
  arXiv  Detail & Related papers  (2025-07-22T18:37:33Z)
• An Agentic System for Rare Disease Diagnosis with Traceable Reasoning [58.78045864541539]
  We introduce DeepRare, the first rare disease diagnosis agentic system powered by a large language model (LLM)<n>DeepRare generates ranked diagnostic hypotheses for rare diseases, each accompanied by a transparent chain of reasoning.<n>The system demonstrates exceptional diagnostic performance among 2,919 diseases, achieving 100% accuracy for 1013 diseases.
  arXiv  Detail & Related papers  (2025-06-25T13:42:26Z)
• Quantifying the Reasoning Abilities of LLMs on Real-world Clinical Cases [48.87360916431396]
  We introduce MedR-Bench, a benchmarking dataset of 1,453 structured patient cases, annotated with reasoning references.<n>We propose a framework encompassing three critical examination recommendation, diagnostic decision-making, and treatment planning, simulating the entire patient care journey.<n>Using this benchmark, we evaluate five state-of-the-art reasoning LLMs, including DeepSeek-R1, OpenAI-o3-mini, and Gemini-2.0-Flash Thinking, etc.
  arXiv  Detail & Related papers  (2025-03-06T18:35:39Z)
• A Scalable Approach to Benchmarking the In-Conversation Differential Diagnostic Accuracy of a Health AI [0.0]
  This study introduces a scalable benchmarking methodology for assessing health AI systems.<n>Our methodology employs 400 validated clinical vignettes across 14 medical specialties, using AI-powered patient actors to simulate realistic clinical interactions.<n>August achieved a top-one diagnostic accuracy of 81.8% (327/400 cases) and a top-two accuracy of 85.0% (340/400 cases), significantly outperforming traditional symptom checkers.
  arXiv  Detail & Related papers  (2024-12-17T05:02:33Z)
• Medchain: Bridging the Gap Between LLM Agents and Clinical Practice through Interactive Sequential Benchmarking [58.25862290294702]
  We present MedChain, a dataset of 12,163 clinical cases that covers five key stages of clinical workflow.<n>We also propose MedChain-Agent, an AI system that integrates a feedback mechanism and a MCase-RAG module to learn from previous cases and adapt its responses.
  arXiv  Detail & Related papers  (2024-12-02T15:25:02Z)
• AI Workflow, External Validation, and Development in Eye Disease Diagnosis [5.940140611616894]
  AI shows promise in diagnosis accuracy but faces real-world application issues due to insufficient validation in clinical and diverse populations.<n>This study addresses gaps in medical AI downstream accountability through a case study on age-related macular degeneration (AMD) diagnosis and classification severity.
  arXiv  Detail & Related papers  (2024-09-23T15:01:09Z)
• TrialBench: Multi-Modal Artificial Intelligence-Ready Clinical Trial Datasets [54.98321887435557]
  This paper presents a suite of 23 meticulously curated AI-ready datasets covering multi-modal input features and 8 crucial prediction challenges in clinical trial design.<n>We provide basic validation methods for each task to ensure the datasets' usability and reliability.<n>We anticipate that the availability of such open-access datasets will catalyze the development of advanced AI approaches for clinical trial design.
  arXiv  Detail & Related papers  (2024-06-30T09:13:10Z)
• AI Hospital: Benchmarking Large Language Models in a Multi-agent Medical Interaction Simulator [69.51568871044454]
  We introduce textbfAI Hospital, a framework simulating dynamic medical interactions between emphDoctor as player and NPCs. This setup allows for realistic assessments of LLMs in clinical scenarios. We develop the Multi-View Medical Evaluation benchmark, utilizing high-quality Chinese medical records and NPCs.
  arXiv  Detail & Related papers  (2024-02-15T06:46:48Z)
• Beyond Direct Diagnosis: LLM-based Multi-Specialist Agent Consultation for Automatic Diagnosis [30.943705201552643]
  We propose a framework to model the diagnosis process in the real world by adaptively fusing probability distributions of agents over potential diseases. Our approach requires significantly less parameter updating and training time, enhancing efficiency and practical utility.
  arXiv  Detail & Related papers  (2024-01-29T12:25:30Z)
• Towards Conversational Diagnostic AI [32.84876349808714]
  We introduce AMIE (Articulate Medical Intelligence Explorer), a Large Language Model (LLM) based AI system optimized for diagnostic dialogue. AMIE uses a self-play based simulated environment with automated feedback mechanisms for scaling learning across diverse disease conditions. AMIE demonstrated greater diagnostic accuracy and superior performance on 28 of 32 axes according to specialist physicians and 24 of 26 axes according to patient actors.
  arXiv  Detail & Related papers  (2024-01-11T04:25:06Z)
• Advancing COVID-19 Diagnosis with Privacy-Preserving Collaboration in Artificial Intelligence [79.038671794961]
  We launch the Unified CT-COVID AI Diagnostic Initiative (UCADI), where the AI model can be distributedly trained and independently executed at each host institution. Our study is based on 9,573 chest computed tomography scans (CTs) from 3,336 patients collected from 23 hospitals located in China and the UK.
  arXiv  Detail & Related papers  (2021-11-18T00:43:41Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.