5 key takeaways from a new report on AI, machine learning in radiology

Research firm Reaction Data has published a new report, “Machine Learning in Medical Imaging,” that breaks down what radiologists and other imaging professionals think about AI, machine learning and the future of radiology.

“The primary motivation behind this study is the sheer amount of hype going on in healthcare, specifically in radiology and imaging, around AI—deep learning and machine learning,” the report’s authors wrote. “In essence, the machine learning buzz is, quite literally, through the roof.”

Reaction Data received input from more than 130 industry professionals. While 45 percent of respondents were directors of radiology, another 20 percent were radiologists and 9 percent were imaging directors. Radiology managers, PACS administrators and others also participated.

These are 5 key takeaways from their analysis:

1. The industry understands the importance of AI and machine learning.

Industry professionals—especially those enjoying a lot of success—are often in denial about certain changes. After all, why adapt to something new when you can just ignore it? According to the report, this is not the case when it comes to radiology and machine learning.

Respondents were asked to grade the importance of machine learning using a scale of one to seven, with seven being “extremely important.” One out of four respondents chose a seven, and another 59 percent chose a four, five or six.

“Only 16 percent of respondents think machine learning just isn’t that important, with the vast majority viewing machine learning technology as being either important or extremely important in medical imaging,” the authors wrote.

2. Even those who aren’t very familiar with machine learning are buying into the hype.

Respondents were asked to grade how familiar they are with machine learning on the same one-to-seven scale, with seven being “extremely familiar.” While 28 percent of respondents chose a five, the most common response by a significant margin, 17 percent chose just a one or two. Another 16 percent chose a three, and 15 percent chose a four.

As the report pointed out, this seems to show the power of the buzz surrounding AI and machine learning; even those who don’t understand the details think it will have a huge impact. “We may be dealing with a mild case of FOMO (fear of missing out),” the authors noted.

3. You think machine learning is big today? Give it a year or two and it'll be much bigger.

Twenty-seven percent of respondents are prepared to adopt machine learning in the next 1-2 years, and another 11 percent are less than 12 months away. Twenty-three percent of respondents, meanwhile, answered that they “just” adopted machine learning or they’ve been using it “for a while.”

4. The No. 1 use case for machine learning right now is breast imaging.

What do imaging professionals use (or plan to use) machine learning for? Sixty-eight percent of respondents answered breast imaging, the highest percentage of any specialty. In addition, 61 percent answered lung imaging and 58 percent answered chest x-rays. Bone imaging, cardiovascular imaging and liver imaging were other common answers.

5. Some organizations aren’t adopting machine learning right now—and they have their reasons.

When reviewing the feedback of those who say they will never utilize machine learning we find their hesitancy to be rather fluid,” the authors wrote. “If their respective organizations become more “forward thinking” and as AI continues to advance there is a strong chance these hold-outs will give in and join their peers in using AI algorithms.”

While 46 percent of respondents who are not adopting machine learning said it’s because they are “unsure of its usefulness right now,” 39 percent of that same group said they “just aren’t that forward thinking.” Finally, 15 percent of respondents who are not adopting machine learning said they simply think humans outperform AI.

Overall, the Reaction Data’s authors believe AI “is here to stay.” They plan to research this again later in the year, and only time will tell how the industry’s thoughts on AI and machine learning change in the next several months.

The full report is available on Reaction Data’s website.