Problem
Research questions and friction points this paper is trying to address.
diagnostic error
patient safety
must-not-miss conditions
clinical reasoning
AI-assisted diagnosis
A safety-oriented hypothetico-deductive framework for AI-assisted differential diagnosis
🤖 AI Summary
This study addresses critical safety concerns in the clinical deployment of large language models (LLMs), particularly their inability to reliably screen for high-risk conditions and provide verifiable diagnostic reasoning. To this end, the authors propose AegisDx, a novel framework that introduces, for the first time, a safety-prioritized hypothetico-deductive reasoning mechanism. AegisDx integrates role-specialized LLM components, structured intermediate outputs, a medical evidence retrieval interface, and a validation gating mechanism to jointly ensure systematic exclusion of life-threatening diseases and enable traceable differential diagnosis. Experimental results demonstrate that AegisDx significantly outperforms baseline models in top-3 diagnostic accuracy across multiple medical case collections from peer-reviewed journals, achieves a physician-blind safety rating of 4.55 out of 5, and improves high-risk disease detection by 26 percentage points.
📝 Abstract
Diagnostic error is a major threat to patient safety, yet current large language model (LLM) systems often treat diagnosis as a one-shot prediction task, lacking safeguards against missed high-risk alternatives or rigorous verification of their reasoning. Here, we present AegisDx, a safety-oriented framework for hypothetico-deductive clinical reasoning. AegisDx coordinates specialized LLM components through role-specific contracts, structured intermediate outputs, evidence-retrieval interfaces, and verification gates to generate broad differential diagnoses, enforce explicit screening for dangerous "must-not-miss" conditions, verify reasoning against grounded medical evidence, and structure actionable next steps. We evaluated AegisDx across three layers. On literature-derived case reports from NEJM and JAMA, with GPT-oss-120B as the shared backbone, Top-3 diagnostic accuracy was 59.9% versus 52.1% for the standalone LLM on JAMA cases and 62.7% versus 51.4% on NEJM cases. On cases from Annals of Emergency Medicine, Top-3 accuracy was 85.7% versus 68.6%; against physician-consensus must-not-miss diagnosis sets, AegisDx captured at least one such condition among its top three diagnoses in 78.0% of cases versus 52.0%. In a blinded physician evaluation of 43 real-world emergency department notes from the Yale New Haven Health System compared against GPT-5, AegisDx improved the physician-rated composite safety score from 4.31 to 4.55 on a 5-point scale (adjusted p = 2.1x10^-4), with qualitative gains in must-not-miss identification and reasoning safety. Our findings suggest that engineering diagnostic AI as a safety-oriented reasoning framework, rather than optimizing raw predictive accuracy alone, can provide a safer, more transparent, and clinically meaningful layer of bedside decision support for acute care workflows.
Innovation
Methods, ideas, or system contributions that make the work stand out.
hypothetico-deductive reasoning
safety-oriented AI
differential diagnosis
must-not-miss conditions
structured clinical reasoning
🔎 Similar Papers
No similar papers found.
