Robots Have Learned to Predict Death
CHICAGO – July 1, 2018
Now neural networks can predict a patient's death with two times more accuracy than human doctors— Google had created a tool that could forecast a host of patient outcomes, including how long people may stay in hospitals, their odds of readmission and the chances they will soon die.
For its recent research, the internet giant cut deals with the University of California, San Francisco, and the University of Chicago for 46 billion pieces of anonymous patient data. Google’s AI system created predictive models for each hospital, not one that parses data across the two, a harder problem. A solution for all hospitals would be even more challenging. Google is working to secure new partners for access to more records.
Hospitals, doctors and other health-care providers have been trying for years to better use their stockpiles of electronic health records and other patient data. But current methods of mining health data are costly, cumbersome and time consuming. The neural net gobbled up all this unruly information then spat out predictions. And it did it far faster and more accurately than existing techniques.
From a technological point of view, this is not surprising: AI is great for organizing all the data available for the patient.
Our medicine, with all its sparkling and bizarre devices, remains at a very low-technical level. A human doctor is physically unable to study the full history of the patient or sort out the symptoms of those thousands of diseases from which the patient could suffer.
Robotic systems are doomed to win the primary diagnosis chase with humans for the same reasons that online search systems won against the human librarians and archivists digging through their paper forms and files.
Of course, doctors will not disappear from our life because of the widespread introduction of medical robotics. But they will have to get used to a new reality and perhaps ask a robot’s recommendations for a treatment.