One-third of U.S. facilities are using AI in aiding imaging studies

A third of American hospitals and imaging centers are already implementing artificial intelligence, machine learning or deep learning in radiology.

About another third plan on adopting it within the next two years, according to respondents to a survey, the results of which were recently released by an industry research organization.

The survey of 207 imaging leaders and professionals, conducted by healthcare data and analytics firm Definitive Healthcare from October through December, found that the use of AI was seen as “highly valuable” in enhancing patient care and business operations. The survey was Definitive Healthcare’s first study of AI use in imaging.


A majority of respondents (57.3 percent) say they believe that AI will have the greatest impact on patient care by improving the accuracy of medical diagnoses and clinical outcomes. AI technologies were also seen as offering operational improvements, faster diagnoses, early detection and improvements in the quality of existing imaging techniques, such as MRIs.

Fewer than 1 percent stated that AI would provide long-term cost savings.

Imaging centers use AI slightly (2.8 percentage points) more than hospitals, possibly because their primary focus is on diagnostics and because they may be more willing to risk using a new technology, according to the study.

Current users of AI in radiology overwhelmingly (93 percent) use it in computer-aided image detection for diseases.

Respondents say they use AI very little for other applications; 27 percent use it for process or workflow improvement, and 16 percent use it for technological monitoring or equipment maintenance. Only 15 percent use AI technology for computer aided image detection for fractures or musculoskeletal injuries, and only 13 percent use it for care guideline consultations or to suggest care options.

The biggest challenge to adopting and using AI cited is its cost (by 54.7 percent of respondents), followed by “lack of strategic direction” (35.2 percent) and lack of technical expertise (33.1 percent). Some 19 percent of respondents cited lack of leadership buy-in as a barrier to AI adoption.

The global AI in medical imaging market is expected to rise from $21.5 billion in 2018 to a projected $264.9 billion by 2026, according to Data Bridge Market Research’s 2019 report.

“The widespread interest in these technologies, successful use cases, and continued research and development will likely result in rapid growth and implementation for AI [in relation to imaging] in the near future—particularly as these technologies become more widely-accessible over the next few years,” said Jason Krantz, CEO of Definitive Healthcare.