[ad_1]
However, it is accurate just over half the time, which raises questions about whether and how it could be used in real-life conditions.
The news: Deepmind, Google's artificial intelligence unit, has created an in-depth learning system that badyzes a person's medical records, including information such as vital signs and blood test results, and then predicts the risk of acute kidney injury (AKI). In a paper published in the journal Nature The researchers showed yesterday that they were able to predict 55.8% of cases up to 48 hours before their onset. For more severe kidney damage, such as those requiring dialysis at a later stage, the accuracy was closer to 90%.
The promise: AKI contributes to nearly 300,000 deaths in the United States each year, and affects one in five patients admitted to the hospital in America for serious care. However, if it is caught and treated early, it can be reversed. That's why this system is so promising: it could potentially save many lives.
However: It was built with the help of US Department of Veterans Records, a set of training data consisting of 94% of men. Much more testing is needed to confirm its utility in the general population. There were also a lot of false positives.
DeepMind plans to integrate the algorithm into its Streams application, which helps doctors at the Royal Free Hospital in London identify patients at risk of developing AKI. Only in this way, after further tests, can we really say how useful it will be.
To learn more about the world of AI, sign up right here to our Weekly newsletter on the AI, the algorithm.
Source link
