Google’s DeepMind turns its virtual hand to predicting when someone might die

AI can’t quite help us live forever, but it could help aid medical professionals in finding the causes of unexpected patient deaths.

Google's DeepMind turns its virtual hand to predicting when someone might die

Working with the US Veterans Administration, Google’s sister-company and AI behemoth DeepMind is attempting to learn how to predict the changes in a patient’s condition that, when left unchecked, could lead to death.

READ NEXT: If you can smell “death”, you may be at greater risk of depression

DeepMind’s partnership means its AI is checking the historical medical records of around 700,000 US veterans to spot “human error” around their treatment and is just the latest example of AI picking up the smaller tasks in medical care to ensure humans live more comfortable lives. Only last month did we hear about how AI was going to be rolled out in hospitals to detect heart disease and cancer, and before that, it was said it could be used to detect anomalies in colonoscopy examinations.

Using anonymised medical data taken from US veterans, DeepMind analysed the data in a bid to figure out what causes unexpected patient deterioration in a bid to spot it ahead of time. Traditionally, the role of monitoring patient conditions falls to nurses but with an increasing strain on medical resources, it’s hard to expect nurses and doctors to watch patients all the time outside of doing their standard medical rounds.

READ NEXT: 10 things you need to know about the future of AI

Even with an array of sensors set up designed to alert medical professionals to a patient’s state, it’s still impossible to expect anyone to be there to respond to everyone at all times of the day.

deepmind_ai_patent_death_-_future_doctorThe doctors and nurses of the future

Unlike a nurse or doctor, an AI doesn’t get tired and can be omnipresent, watching over every patient it’s assigned to, all at the same time. It may not be the be-all, end-all of unexpected patient deaths in hospitals but it certainly goes a fair way towards reducing the death count in US hospitals as human error is, according to Johns Hopkins Medicine research, the third-leading cause of death in the US.

Deepmind isn’t diving into every single sign of patient deterioration listed in in the anonymous medical records. Instead, it’s focusing on Acute Kidney Injury (AKI), a leading condition associated with patient deterioration. By focusing on just this one area, DeepMind is laying the groundwork for more complex and far-reaching issues.

READ NEXT: AI vs machine learning, what’s the difference?

“This is a complex challenge, because predicting AKI is far from easy,” explains DeepMind’s blog post. “Not only is the onset of AKI sudden and often asymptomatic, but the risk factors associated with it are commonplace throughout hospitals. AKI can also strike people of any age, and frequently occurs following routine procedures and operations like a hip replacement.”

If DeepMind’s plans play out as they expect, the implication upon the quality of life of patients, along with humanity’s general life expectancy, would be huge. We’d live longer, fight fitter and catch early signs of anything potentially life-threatening where it’s at a point we can deal with it.

Still, don’t go expecting something as miraculous as the radiology machines from Elysium. We’re a long way off that depressingly dark future…

Disclaimer: Some pages on this site may include an affiliate link. This does not effect our editorial in any way.

Todays Highlights
How to See Google Search History
how to download photos from google photos