AI vs. Humans: Microsoft's Neural Net Beats Doctors In diagnosing Rare Diseases

Science & Technology Technology

Posted by AI on 2025-07-01 16:37:11 | Last Updated by AI on 2025-07-01 18:04:26

Share: Facebook | Twitter | Whatsapp | Linkedin Visits: 0


AI vs. Humans: Microsoft's Neural Net Beats Doctors In diagnosing Rare Diseases

Microsoft has developed an artificial intelligence model that boasts quadruple the success rate of human doctors in diagnosing rare medical conditions that defy typical symptoms and tests. The AI a neural network was tested against the diagnoses of six rare disease experts, and it significantly out-performed the humans in both accuracy and speed.

This is sensational news for those suffering from mysterious ailments, and a landmark step forward for Microsoft in the growing role of AI in healthcare. "It's very unusual for a machine to beat humans because these are really hard cases," said one of the Microsoft researchers involved in the study. Indeed, the AI is able to make accurate diagnoses with 70% accuracy, compared to the human experts' 50% a whopping 40% improvement.

While the AI model is a long way from being integrated into hospitals or medical practices, the tech giant has made progress on that front already, having developed a chatbot that acts as a hypochondriac's best friend, and a separate AI that predicts sudden death in patients suffering from heart failure. This latest neural network demonstrates the immense potential for AI in medicine, and with further refinement and development, could be a game-changer for healthcare providers and patients alike.

Search
Categories