Twitter users are optimistic about AI's integration into radiology
Turns out, Twitters users are pretty optimistic about artificial intelligence (AI) in radiology. Also, many users of the social platform seem skeptical that AI could potentially replace radiologists, wrote authors Julia Goldberg, and Andrew Rosenkrantz, MD, from NYU Langone Health in a study published July 23 in Current Problems in Diagnostic Radiology.
Previous studies found Twitter to be the social media of choice for most radiologists—and it continues to be used to promote academic discussions for radiologists. It serves as a public platform allowing for diverse perspectives on AI and radiology outside of the academic sphere, according to the authors.
"Radiologists can more fully engage with the expansion of AI though participating in these wider discussions, even if not working directly on these tools," they wrote. "Participating in social media discussions allows physicians to readily share their perspectives with others, including industry-related individuals who may be actively involved in developing AI applications for radiology."
Goldberg and Rosenkrantz searched Twitter for all tweets mentioning the terms "artificial intelligence" and “radiology" that were posted from November 2016 to October 2017.
They reviewed 605 tweets for the following characteristics:
- Category of user posting the tweet (radiologist, non-radiologist physician, industry- related individual, media/marketing-related individual, radiology practice/facility, radiology-related organization, non-radiology healthcare organization, technology/data organization, healthcare media, technology/data media, individual, other/unknown).
- Geographic location of user posting the tweet (U.S., non-U.S., unable to classify).
- Mention of specific themes within the tweet text (machine/deep learning, specific clinical application, medical society/conference, lecture, academic journal/research, university, industry, start-up/investments, AI replacing radiologists, AI transforming radiology, limitations of AI).
- Stance of the tweet regarding the impact of AI on radiology (favorable, unfavorable, neutral).
- Presence of a working link to an external website.
- Category of link source (radiology media, healthcare media, technology/data media, mainstream media, non-media healthcare-related organization, technology/data organization, social media/personal blog, other/unknown).
- Mention of specific themes related to AI within the linked website (efficiency improvements, ethical issues, legal/regulatory issues, other limitations/challenges).
- Presence of content in linked websites specifically related to the topic of AI and radiology.
- Stance of linked websites regarding the issues of AI replacing radiologists and the impact of AI on radiology.
Some 407 users (22.6 percent industry-related individuals and 9.3 percent radiologists) generated the tweets and linked to 2016 unique websites. Of these users, 42.5 percent were from the U.S., and 17.2 percent of tweets mentioned machine/deep learning.
Roughly a fourth of tweets expressed optimism regarding the impact of AI on radiology, while 75 percent had a neutral stance. No tweets were unfavorable of the idea. The authors also found that 88 percent of linked websites in the tweets learned towards AI being positive for the field of radiology, and none were explicitly negative.
Of the 47.3 percent of websites linked that mentioned the issue of AI replacing radiologists, 77.5 percent were against AI replacing radiologists, according to the authors.
A recent graduate from Dominican University (IL) with a bachelor’s in journalism, Melissa joined TriMed’s Chicago team in 2017 covering all aspects of health imaging. She’s a fan of singing and playing guitar, elephants, a good cup of tea, and her golden retriever Cooper.