US leads the world in publishing AI-based radiology research
The United States is a global leader in publishing AI-based radiology research, according to new findings published in the American Journal of Roentgenology.
“In the field of medicine, radiology in particular lends itself to AI research because of its large digital data sets,” wrote Elizabeth West, New York Presbyterian Hospital in New York City, and colleagues. “In response, the radiology community has largely embraced AI research, as has been shown by the growing number of publications focusing on such research and the attention it has been given at large radiology society meetings.”
West et al. explored a “comprehensive central database” for radiology-specific research related to AI published from 2000 to 2018 using various search terms related to both AI and radiology. Both linear and nonlinear regression analyses were performed to track how different contributing countries compared to one another. More than 8,800 research articles were included in the analysis.
Overall, the United States has dominated publications related AI and radiology. From 2000 to 2018, the country has accounted for anywhere from 35 to 50% of all AI-based radiology research being published around the world. China was responsible for 18% of all AI-based radiology research being published in 2018, showing the country’s rise in this area.
“China's ability to exponentially increase productivity is likely due to the country's unique research infrastructure,” the authors wrote. “The availability of large centralized data and rapid implementation across commercial industries have already helped the nation become very productive in AI research in a short period. In addition, Chinese government directives and funding for the advancement of AI have generated an incredible mobilization in research and development among Chinese researchers.”
West et al. also found that the United States (16.5%), China (3.6%), the UK (1.5%) and Canada (0.6%) were the four top funding sources for these published pieces. In addition, neuroradiology was the subspecialty that focused on these publications the most.
“The neuroradiology subspecialty produced the most AI publications in this study,” the authors wrote. “This is not surprising, given that neuroradiology is a unique subspecialty of acuity, where ‘time is brain’ for stroke evaluation demands fast and accurate diagnoses suitable for AI applications. In addition to tumor assessment, a common AI application shared with other subspecialties, other neurologic disease processes, including psychiatric disorders, traumatic brain injury, demyelinating diseases (multiple sclerosis), and dementia (Alzheimer disease), are well matched for AI applications.”
Michael has more than 18 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.