Scanning The Future, Radiologists See Their Jobs At Risk
These days, a radiologist at UCSF will go through anywhere from 20 to 100 scans a day, and each scan can have thousands of images to review.
Source: Lauren Silverman
In health care, you could say radiologists have typically had a pretty sweet deal. They make, on average, around $400,000 a year — nearly double what a family doctor makes — and often have less grueling hours. But if you talk with radiologists in training at the University of California, San Francisco, it quickly becomes clear that the once-certain golden path is no longer so secure.
"The biggest concern is that we could be replaced by machines," says Phelps Kelley, a fourth-year radiology fellow. He's sitting inside a dimly lit reading room, looking at digital images from the CT scan of a patient's chest, trying to figure out why he's short of breath.
Because MRI and CT scans are now routine procedures and all the data can be stored digitally, the number of images radiologists have to assess has risen dramatically. These days, a radiologist at UCSF will go through anywhere from 20 to 100 scans a day, and each scan can have thousands of images to review.
"Radiology has become commoditized over the years," Kelley says. "People don't want interaction with a radiologist, they just want a piece of paper that says what the CT shows."
'Computers are awfully good at seeing patterns'
That basic analysis is something he predicts computers will be able to do.
Dr. Bob Wachter, an internist at UCSF and author of The Digital Doctor, says radiology is particularly amenable to takeover by artificial intelligence like machine learning.
"Radiology, at its core, is now a human being, based on learning and his or her own experience, looking at a collection of digital dots and a digital pattern and saying 'That pattern looks like cancer or looks like tuberculosis or looks like pneumonia,' " he says. "Computers are awfully good at seeing patterns."
Just think about how Facebook software can identify your face in a group photo, or Google's can recognize a stop sign. Big tech companies are betting the same machine learning process — training a computer by feeding it thousands of images — could make it possible for an algorithm to diagnose heart disease or strokes faster and cheaper than a human can.
UCSF radiologist Dr. Marc Kohli says there is plenty of angst among radiologists today.
"You can't walk through any of our meetings without hearing people talk about machine learning," Kohli says.
Both Kohli and his colleague Dr. John Mongan are researching ways to use artificial intelligence in radiology. As part of a UCSF collaboration with GE, Mongan is helping teach machines to distinguish between normal and abnormal chest X-rays so doctors can prioritize patients with life-threatening conditions. He says the people most fearful about AI understand the least about it. From his office just north of Silicon Valley, he compares the climate to that of the dot-com bubble.
"People were sure about the way things were going to go," Mongan says. "Webvan had billions of dollars and was going to put all the groceries out of business. There's still a Safeway half a mile from my house. But at the same time, it wasn't all hype."
'You need them working together'
The reality is this: dozens of companies, including IBM, Google and GE, are racing to develop formulas that could one day make diagnoses from medical images. It's not an easy task: to write the complex problem-solving formulas, developers need access to a tremendous amount of health data.
Health care companies like vRad, which has radiologists analyzing 7 million scans a year, provide data to partners that develop medical algorithms.
The data has been used to "create algorithms to detect the risk of acute strokes and hemorrhages" and help off-site radiologists prioritize their work, says Dr. Benjamin Strong, chief medical officer at vRad.
Zebra Medical Vision, an Israeli company, provides algorithms to hospitals across the U.S. that help radiologists predict disease. Chief Medical Officer Eldad Elnekave says computers can detect diseases from images better than humans because they can multitask — say, look for appendicitis while also checking for low bone density.
"The radiologist can't make 30 diagnoses for every study. But the evidence is there, the information is in the pixels," Elnekave says.
Still, UCSF's Mongan isn't worried about losing his job.
"When we're talking about the machines doing things radiologists can't do, we're not talking about a machine where you can just drop an MRI in it and walk away and the answer gets spit out better than a radiologist," he says. "A CT does things better than a radiologist. But that CT scanner by itself doesn't do much good. You need them working together."
In the short term, Mongan is excited algorithms could help him prioritize patients and make sure he doesn't miss something. Long term, he says radiologists will spend less time looking at images and more time selecting algorithms and interpreting results.
Kohli says in addition to embracing artificial intelligence, radiologists need to make themselves more visible by coming out of those dimly lit reading rooms.
"We're largely hidden from the patients," Kohli says. "We're nearly completely invisible, with the exception of my name shows up on a bill, which is a problem."
Wachter believes increasing collaboration between radiologists and doctors is also critical.
"At UCSF, we're having conversations about [radiologists] coming out of their room and working with us. The more they can become real consultants, I think that will help," he says.
Kelley, the radiology fellow, says young radiologists who don't shy away from AI will have a far more certain future. His analogy? Uber and the taxi business.
"If the taxi industry had invested in ride-hailing apps maybe they wouldn't be going out of business and Uber wouldn't be taking them over," Kelley says. "So if we can actually own [AI], then we can maybe benefit from it and not be wiped out by it."
At least for now, Kelley offers what a computer can't — a diagnosis with a face-to-face explanation.
Copyright 2017 KERA. To see more, visit KERA.
ARI SHAPIRO, HOST:
It's Labor Day, so we're looking at jobs on this week's All Tech Considered.
(SOUNDBITE OF MUSIC)
SHAPIRO: Today we're kicking off a new series that looks at how advances in artificial intelligence are changing our work. It's called Is My Job Safe? We'll look at specific industries where jobs might be disappearing or changing. To begin, we're going to look at which parts of the workforce might be relatively safe from the robots. We're joined by Erik Brynjolfsson. He directs the MIT Initiative on the Digital Economy. Welcome to the program.
ERIK BRYNJOLFSSON: Good to be here.
SHAPIRO: Back in 2004, Researchers at MIT and Harvard published a list of professions that they felt were most and least likely to undergo automation. And one example they gave of a job that could not possibly be automated in the future was truck driving.
SHAPIRO: And today automated vehicles are being tested on the roads. Already the job of truck driving could be completely automated. So your job is to try to predict which jobs will be automated in the future. But I wonder, are humans really able to make these kinds of predictions? The evidence seems to be that we're not very good at it.
BRYNJOLFSSON: It's definitely not easy. There's constantly new innovations coming along, as there should be. And so we have to update our insights from time to time.
SHAPIRO: Well, with that as a caveat, how much of the U.S. workforce would you say is at risk of automation in the coming decades? Are we talking about, like, 10 percent, 50 percent, 80 percent?
BRYNJOLFSSON: Well, I've got to give you some perspective. There's constantly automation of huge chunks of the workforce. And there's new jobs being created and old jobs being automated. And that's going to happen in the next 10 years. I wouldn't be surprised if 50 percent or more of the existing jobs had to change drastically or were eliminated. And hopefully another 50 percent of new jobs will be created at the same time.
SHAPIRO: What do you see as the sector of the workforce that is least likely to change or least likely to disappear?
BRYNJOLFSSON: Well, there are three big categories that machines are really bad at. They've made tremendous advances, but they're bad at first off doing creative work. Whether you're an entrepreneur or a scientist or a novelist, I think you're in pretty good shape doing that long-range creativity. The second big category is interpersonal skills and emotional intelligence, people who are coaches or salespeople or negotiators or caregivers. And the third one is actually manual dexterity and physical mobility. Machines have a hard time doing simple things like picking up a nickel or walking up stairs or clearing a table.
And so jobs that depend on that will also be safe for a while. And I think the right way to think about it is not so much looking at jobs, but looking at tasks 'cause almost every job has parts of them that are in one of those three categories, or maybe all three, and other parts that will be affected or even automated.
SHAPIRO: It's interesting 'cause when I think about how that translates to education, there's been such an emphasis on science and technology education. But it sounds like you're saying one of the sectors that's likely to be safest is sort of creative work that would suggest liberal arts education.
BRYNJOLFSSON: Absolutely. In fact, I think there's probably no better time in history to be somebody with some real creative insights. And then the technology helps you leverage that to millions or billions of people. And people who can combine some creativity with an understanding of the digital world are especially well-positioned.
SHAPIRO: Would you say that blue-collar workers are generally more likely to be replaced by robots than white-collar workers? We hear so much about people in manufacturing being replaced by automation.
BRYNJOLFSSON: Well, the truth is most blue-collar work has already been automated. I mean, there's - less than 10 percent of Americans now work in the manufacturing sector. I don't think it's so much of a blue collar-white collar division. The big waves have been more structured work versus less structured work, with more structured work being automated faster and work that involves creativity and interpersonal skills as being more robust in the long run.
SHAPIRO: If people are at the midpoint in their career right now and they want to prepare themselves for the oncoming robot invasion (laughter), what can they do to make it less likely that they will ultimately someday be replaced?
BRYNJOLFSSON: Well, I don't think as a society we're investing enough in education and training and thinking about how to handle this transition. More people should be thinking about the ways we're talking about it right now.
SHAPIRO: Erik Brynjolfsson directs the Initiative on the Digital Economy at MIT, and his latest book is called "Machine, Platform, Crowd." Thanks a lot.