
Foundation 29 and Microsoft created AI for diagnosing rare diseases
The medical neural network from Microsoft called DxGPT may change the approach to diagnosing rare diseases. Microsoft head Satya Nadella says about it: “A great example of what’s possible when a person connects their passion with their purpose to make real change. DxGPT, a tool for diagnosing rare diseases, is already working. Showing how artificial intelligence can be used to truly improve lives.”
Let me tell you more. Behind the development is the non-profit organization Foundation 29. Which has been working relatively long on improving the lives of people with rare and chronic diseases. The operating principle is extremely simple. You enter symptoms, and the neural network gives you several possible diagnoses. The system shows in detail which of your symptoms correspond to each suspected diagnosis, and which do not. You can click on each option and get detailed information.
It’s important to understand that DxGPT is created specifically for doctors, not for patients. The system provides a list of ways to check symptoms, explains why such diagnoses are proposed, allows marking additional symptoms. And gives the doctor a detailed checklist for further actions. It’s a diagnostic tool, not a treatment one.
Several articles confirm that many doctors in Europe are already actively using this neural network in their practice. At the same time, the developers honestly emphasize that DxGPT is still in the development stage. And is intended to support decision-making. Not to replace professional medical judgment. The system makes preliminary diagnoses. It doesn’t analyze images or write prescriptions – it specifically diagnoses. And this is critically important at this stage of artificial intelligence development. They say let the final decisions be made by a human. And it is emphasized that currently the algorithms are not yet perfect enough to fully entrust them with patient treatment.
The service works for free, in 7 languages. And you can use it even without registration.