Artificial Intelligence and Radiology: A Social Media Perspective.

Area of Science:

• Medical informatics
• Digital health
• Radiology research

Background:

• Artificial intelligence (AI) is rapidly advancing, with significant implications for medical fields.
• Understanding public and professional perceptions of AI in radiology is crucial for its adoption.

Purpose of the Study:

• To analyze public perspectives on artificial intelligence (AI) in radiology using Twitter data.
• To characterize the sentiment and key themes in online discussions about AI and radiology.

Main Methods:

• A systematic search of Twitter for tweets containing "artificial intelligence" and "radiology" was conducted from November 2016 to October 2017.
• Tweets, user demographics, and linked websites were categorized to analyze content and sentiment.

Main Results:

• Analysis of 605 tweets from 407 unique users revealed predominantly neutral to favorable views on AI in radiology.
• Linked websites overwhelmingly supported AI's positive impact, frequently mentioning improved efficiency, though challenges and job displacement concerns were also noted.
• Industry professionals and US-based users were prominent, with discussions often mentioning machine learning, industry, and medical societies.

Conclusions:

• Online discussions, primarily from Twitter, indicate a positive outlook on AI's transformative potential in radiology.
• While acknowledging implementation challenges, the discourse generally opposes the idea of AI replacing radiologists.
• Increased engagement from radiologists in these social media conversations is recommended to shape the narrative and facilitate AI integration.