Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
August 14, 2025Open Access

AI vs Human Performance in Conversational Hospital-Based Neurological Diagnosis

View Full Paper
Ask AI
Bookmark
Share

Authors

MSMoran SorkaAGAlon GorenshteinHAHillel Abramovitch

Discussion

Loading...

Member takes

Overview

Evaluation shows AI achieves higher diagnostic accuracy and lower costs than human neurologists in clinical settings.

Key Points

  • AI systems, particularly a multi-agent model, achieved better diagnostic accuracy than human neurologists, with 81% accuracy for humans and 94% for AI.
  • Gregory, the AI model, provided faster diagnoses at an average cost of $1,423, significantly less than human neurologists' average of $3,041.
  • The study involved assessing 14 neurologists and multiple state-of-the-art large language models across 16 clinical cases for accuracy, costs, and time.
  • Findings highlight how AI enhances diagnostic efficiency, suggesting a need for integrating conversational AI in clinical practices.

Cite This Study

Sorka et al. (2025) studied this question.

synapsesocial.com/papers/68af3e3ccf1dd9ea359eaf10https://doi.org/10.1101/2025.08.13.25333529
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1A Multi-AI Agent Framework for Interactive Neurosurgical Education and Evaluation: From Vignettes to Virtual Conversations2025
  2. 2Toward the Autonomous AI Doctor: Quantitative Benchmarking of an Autonomous Agentic AI Versus Board-Certified Clinicians in a Real World Setting2025 · 1 citations
  3. 3Toward the Autonomous AI Doctor: Quantitative Benchmarking of an Autonomous Agentic AI Versus Board-Certified Clinicians in a Real World Setting2025 · 10 citations
  4. 4The Cognitive Safety Net: Comparing Human and AI Diagnostic Reasoning during Complex Clinical Situations2025
  5. 5Evaluating Large Language Model Diagnostic Performance on JAMA Clinical Challenges via a Multi-Agent Conversational Framework2025