Using the power of artificial intelligence (AI) doctors could predict when their patients might be knocking on death’s door. This isn't merely some grim dystopian prop, the researchers hope it could be used to slash the surprisingly high number of unexpected deaths in the US.
Excel Medical, a medical tech company in Florida, has recently been boasting about its new WAVE Clinical Platform, an algorithm that can accurately predict whether medical patients could be at risk of a sudden, unexpected death.
It consists of an integrated system of hospital workstations and digital medical records that includes real-time data on their physiology, past medical history, family history, medications, age, etc. With all this information, the AI can automatically calculate the risk of patient deterioration up to six hours in advance of when a doctor might notice. If it detects anything suspicious, the system can inform an on-call doctor through a smartphone app.
It has also just become the first AI platform of its kind to be cleared by the US Food and Drug Administration (FDA). This decision was based on a series of studies at the University of Pittsburgh Medical Center that showed the AI platform could prevent unexpected deaths in hospitals. Another more recent study, using similar technology by Stanford University, outlined on the preprint server arXiv, explains how a deep-learning algorithm can correctly predict an otherwise-unexpected death in 90 percent of cases.
“Everything we do as an organization aligns toward and supports the goal of eradicating unexpected deaths in hospitals,” Lance Burton, General Manager of Excel Medical, said in a statement. “People may say zero unexpected deaths is unattainable. We say anything other than zero is unconscionable.”
Currently, the plan is to implement this tech in hospitals, but Burton told Healthcare Analytics News that the company hopes to eventually make the AI wearable devices usable for people at home.
Biomedicine is one of the most exciting and promising applications for AI around. Stanford University also developed a new AI that is as accurate as doctors at identifying skin cancer from images, and hopes to get everyday smartphones to carry it.