Microsoft AI Publishes Medical Diagnostic AI Research in Science, Announces Mayo Clinic and Beth Israel Deaconess Partnerships
Microsoft AI published research in Science on its MAI-DxO diagnostic AI system, which demonstrated 83% diagnostic accuracy on complex New England Journal of Medicine cases compared to 36% for physicians using reference resources and 22% without them. On typical emergency department presentations from MIMIC-IV-ED, MAI-DxO achieved 80% accuracy versus 47% for physicians, and the system performs diagnostic tasks at approximately one-half to one-third the cost. The company also announced partnerships with Mayo Clini...
Key takeaways
MAI-DxO achieved 83% diagnostic accuracy on 304 complex NEJM cases versus 36% accuracy for physicians with resources and 22% without
On MIMIC-IV-ED emergency department cases, MAI-DxO achieved 80% accuracy versus 47% for physicians—1.7 times higher accuracy
MAI-DxO delivers diagnoses at approximately half to one-third the cost of comparative systems while maintaining high accuracy
Microsoft AI introduced SDBench (Sequential Diagnosis Benchmark) with 304 complex NEJM cases and MIMIC-IV-ED typical ED presentations
The approach uses multiple AI models operating like a panel of virtual physicians proposing diagnoses, selecting tests, and debating findings over several rounds
Performance gains persisted when evaluated with newer models including GPT-5.5 without changing the orchestration method
Microsoft AI announced collaboration with Mayo Clinic to combine frontier AI, clinical expertise, and multimodal data
Key metrics
Primary source
TickerStar generated this brief from an official company, investor-relations, earnings-call, or regulatory source. Review the primary material before making an investment decision.
Open Investor relationsFollow MSFT in real time
Get material earnings, filing, IR, and earnings-call alerts by email.