Microsoft AI Model Outperforms Physicians in Diagnostic Accuracy in Simulations

A new report indicates that Microsoft's MAI-DxO, an AI model, successfully outperformed human physicians in diagnostic accuracy during controlled simulations of medical cases. The model achieved 83% accuracy on New England Journal of Medicine cases compared to 36% for physicians with references, and 80% versus 47% on emergency department cases. This research, published in Science, highlights the potential for AI to lower technical barriers to sophisticated medical diagnosis, though the simulations were text-only and did not include benign cases, imaging, or specialists.