Quantcast
Channel: Artificial Intelligence
Viewing all articles
Browse latest Browse all 1160

The head of healthcare at Microsoft lays out the 3 ways AI will actually transform healthcare

$
0
0

microsoft research peter lee

  • Peter Lee, the corporate vice president of Microsoft Healthcare, has a pretty straightforward thesis when it comes to tech's involvement in healthcare.
  • "For me, it's fundamentally a question of what right do we have as a company like Microsoft to be participating in healthcare," Lee said.
  • With that in mind, he sees three areas in healthcare where artificial intelligence can benefit the work being done.

DAVOS, Switzerland — For Peter Lee, the bar for big tech players like Microsoft to be in the healthcare business has to be really high.

"For me, it's fundamentally a question of what right do we have as a company like Microsoft to be participating in healthcare," Lee, the corporate vice president of Microsoft Healthcare, told Business Insider's executive editor Matt Turner at the World Economic Forum's annual meeting in Davos, Switzerland.

Microsoft for its part is using its cloud-computing service, Microsoft Azure, as a place where healthcare companies can store all the health data they're gathering, ranging from X-rays and other types of medical images to individuals' genetic profiles.

Read more:Microsoft's head of healthcare thought it was a 'career-ending move' when Satya Nadella offered him the job. Here's why he says he's now completely sucked in.

Microsoft could also use its technology to crunch through that data and use it to find new solutions with artificial intelligence. When asked what the opportunity was to use AI in healthcare, Lee replied, "It's sort of like saying, what's so great about sunshine?"

In the near term, however, Lee said there are three areas in particular where AI could come in handy.

Translating early signs of disease

The first way is through using machine learning to sift through information about proteins and the immune system. To that end, Microsoft has partnered with Seattle-based Adaptive Biotechnologies.

The partnership is centered on mapping out the immune system, with the hope of finding new ways to diagnose and treat diseases. To do that, Adaptive's producing more than 1 trillion points of data that the company and Microsoft can use to train machine-learning algorithms.

Ultimately, the hope is to use that data to inform early diagnostic tests for cancer or autoimmune conditions like multiple sclerosis or celiac disease, as well as infectious disease. Lee said the partnership is particularly focused on Lyme disease.

Read more:How Microsoft's top scientists have built a big business in hacking healthcare — and helped a lot of people along the way

To sift through that information, Microsoft's using the same skills as applying machine learning to translate languages.

"It just tickles the computer scientists that we're using the same code that we use for Skype Translator," Lee said.

Aiding the doctor-patient experience

A nearer term use of AI in healthcare, Lee said, is using it to make the relationship between doctors and patients better.

Right now, doctors have a lot of information they need to write down during a patient visit. Companies are building Alexa-like assistants to help fill out paperwork on behalf of doctors.

That's one way AI could cut down on the workload: by using the language and observational tools AI has to automate the process and catching errors the doctors might make when filling out information for insurance claims.

Lee said that's going to be a competitive area of interest, with Microsoft's competitors like Amazon or IBM investing heavily in it.

'Nuts and bolts'

On the less exciting end of the spectrum, AI could be used effectively to help with what seems like the basics, such as aggregating health data and presenting it in a way that preserves patient privacy. Right now, that's not as simple as it might seem.

"It's just not obvious how to take a gigantic amount of, let's say, clinical notes out of the EMR system and ensure that data about you and your identity aren't revealed inadvertently," Lee said.

Join the conversation about this story »

NOW WATCH: An exercise scientist reveals exactly how long you need to work out to get in great shape


Viewing all articles
Browse latest Browse all 1160

Trending Articles