A Google-based AI can determine when patients in the hospital will die with an accuracy of about 95 %, reports independent.co.uk.
The chance to learn how long you may have left on this mortal coil is definitely not for everyone, but a prediction of someone’s mortality can provide benefits for doctors looking after hospital patients, ranging from improved end-of-life care to ensuring that loved ones are well informed in advance.
Google’s Medical Brain artificial intelligence is at the forefront of this technology, which determines how long a patient has been left to live by examining every small detail of their medical record. In a recent study, the predictive algorithm of the system tried to determine when patients would die in two separate hospitals-achieving impressive 93 and 95 percent accuracy rates.
“This was significantly more accurate than the traditional predictive model,” the scientists wrote. “These models outperformed traditional, clinically-used predictive models in all cases.”
“We believe that this approach can be used to create accurate and scalable predictions for a variety of clinical scenarios.”
While the technology is undoubtedly impressive, critics have raised concerns about privacy regarding access to personal information by the system. For example, back in 2013, Google’s DeepMind AI was granted access to 1,6 million NHS patients medical records.
The fact that in the not too distant future, Google AI and others like it will know more about us than we know about ourselves is certainly not beyond possibilities.
