(Press-News.org) OAK BROOK, Ill. – When making diagnostic decisions, radiologists and other physicians may rely too much on artificial intelligence (AI) when it points out a specific area of interest in an X-ray, according to a study published today in Radiology, a journal of the Radiological Society of North America (RSNA).
“As of 2022, 190 radiology AI software programs were approved by the U.S. Food and Drug Administration,” said one of the study’s senior authors, Paul H. Yi, M.D., director of intelligent imaging informatics and associate member in the Department of Radiology at St. Jude Children’s Research Hospital in Memphis, Tennessee. “However, a gap between AI proof-of-concept and its real-world clinical use has emerged. To bridge this gap, fostering appropriate trust in AI advice is paramount.”
In the multi-site, prospective study, 220 radiologists and internal medicine/emergency medicine physicians (132 radiologists) read chest X-rays alongside AI advice. Each physician was tasked with evaluating eight chest X-ray cases alongside suggestions from a simulated AI assistant with diagnostic performance comparable to that of experts in the field. The clinical vignettes offered frontal and, if available, corresponding lateral chest X-ray images obtained from Beth Israel Deaconess Hospital in Boston via the open-source MIMI Chest X-Ray Database. A panel of radiologists selected the set of cases that simulated real-world clinical practice.
For each case, participants were presented with the patient’s clinical history, the AI advice and X-ray images. AI provided either a correct or incorrect diagnosis with local or global explanations. In a local explanation, AI highlights parts of the image deemed most important. For global explanations, AI provides similar images from previous cases to show how it arrived at its diagnosis.
“These local explanations directly guide the physician to the area of concern in real-time,” Dr. Yi said. “In our study, the AI literally put a box around areas of pneumonia or other abnormalities.”
The reviewers could accept, modify or reject the AI suggestions. They were also asked to report their confidence level in the findings and impressions and to rank the usefulness of the AI advice.
Using mixed-effects models, study co-first authors Drew Prinster, M.S., and Amama Mahmood, M.S., computer science Ph.D. students at Johns Hopkins University in Baltimore, led the researchers in analyzing the effects of the experimental variables on diagnostic accuracy, efficiency, physician perception of AI usefulness, and “simple trust” (how quickly a user agreed or disagreed with AI advice). The researchers controlled for factors like user demographics and professional experience.
The results showed that reviewers were more likely to align their diagnostic decision with AI advice and underwent a shorter period of consideration when AI provided local explanations.
“Compared with global AI explanations, local explanations yielded better physician diagnostic accuracy when the AI advice was correct,” Dr. Yi said. “They also increased diagnostic efficiency overall by reducing the time spent considering AI advice.”
When the AI advice was correct, the average diagnostic accuracy among reviewers was 92.8% with local explanations and 85.3% with global explanations. When AI advice was incorrect, physician accuracy was 23.6% with local and 26.1% with global explanations.
“When provided local explanations, both radiologists and non-radiologists in the study tended to trust the AI diagnosis more quickly, regardless of the accuracy of AI advice,” Dr. Yi said.
Study co-senior author, Chien-Ming Huang, Ph.D., John C. Malone Assistant Professor in the Department of Computer Science at Johns Hopkins University, pointed out that this trust in AI could be a double-edged sword because it risks over-reliance or automation bias.
“When we rely too much on whatever the computer tells us, that’s a problem, because AI is not always right,” Dr. Yi said. “I think as radiologists using AI, we need to be aware of these pitfalls and stay mindful of our diagnostic patterns and training.”
Based on the study, Dr. Yi said AI system developers should carefully consider how different forms of AI explanation might impact reliance on AI advice.
“I really think collaboration between industry and health care researchers is key,” he said. “I hope this paper starts a dialog and fruitful future research collaborations.”
###
“Care to Explain? AI Explanation Types Differentially Impact Chest Radiograph Diagnostic Performance and Physician Trust in AI.” Collaborating with Dr. Yi, Dr. Huang, Prinster and Mahmood were Suchi Saria, Ph.D., Jean Jeudy, M.D., and Cheng Ting Lin, M.D.
Radiology is edited by Linda Moy, M.D., New York University, New York, N.Y., and owned and published by the Radiological Society of North America, Inc. (https://pubs.rsna.org/journal/radiology)
RSNA is an association of radiologists, radiation oncologists, medical physicists and related scientists promoting excellence in patient care and health care delivery through education, research and technologic innovation. The Society is based in Oak Brook, Illinois. (RSNA.org)
For patient-friendly information on chest X-rays, visit RadiologyInfo.org.
END
Fukuoka, Japan— Microfluidic technology has become increasingly important in many scientific fields such as regenerative medicine, microelectronics, and environmental science. However, conventional microfabrication techniques face limitations in scale and in the construction of complex networks. These hurdles are compounded when it comes to building more intricate 3D microfluidic networks.
Now, researchers from Kyushu University have developed a new and convenient technique for building such complex 3D microfluidic networks. Their tool? Plants and fungi. The team developed a ‘soil’ medium using nanoparticles of glass (silica) and a cellulose ...
A different mix of fuels with enhanced properties could overcome some of the major barriers to making fusion a more practical energy source, according to a new study.
The proposed approach would still use deuterium and tritium, which are generally accepted as the most promising pair of fuels for fusion energy production. However, the quantum properties of the fuel would be adjusted for peak efficiency using an existing process known as spin polarization. In addition to spin polarizing half the fuels, the percentage of deuterium would be increased from the usual amount of roughly 60% or more.
Models created by scientists at the U.S. Department ...
The American Pediatric Society (APS) is pleased to announce Beth A. Tarini, MD, MS, MBA, as the 2025 Norman J. Siegel New Member Outstanding Science Award recipient for her considerable contributions to pediatric science. The award will be presented to Dr. Tarini during the APS Presidential Plenary at the Pediatric Academic Societies 2025 Meeting in Honolulu, Hawaii, April 24 – April 28.
The award was created in honor of one of the world’s leading nephrologists, Norman J. Siegel, MD, FASN. Dr. Siegel was an outstanding ...
Approximately 16% of the global population, or about 890 million people, suffer from obesity. The World Health Organization (WHO) has recognized obesity as a significant health issue and highlighted its impact on the quality of life and overall health of individuals across the world. In response to this, the World Institute of Kimchi, South Korea, has been publishing a series of articles on the anti-obesity effects of kimchi in international journals. These articles, based on research studies on the subject, indicate that the regular consumption of kimchi, a traditional ...
Researchers delved deep into the regulation of cobalt active sites to enhance the selectivity of propylene to improve scalability and affordability of the production of this important chemical.
Chemical reactions are not always naturally optimized to yield the products in the quantities needed, especially on the scale needed for the amount of industry in the world today. Researchers from East China University of Science and Technology explored the options available to develop a more cost-effective, scalable and straightforward ...
A recent study published in the Journal of Translational Internal Medicine (https://doi.org/10.1515/jtim-2024-0017) reveals significant insights into the broader impact of gastroesophageal reflux disease (GERD) on cardiovascular health. By using a rigorous bidirectional Mendelian randomization (MR) approach, the research provides compelling evidence that GERD—a condition traditionally recognized as a digestive disorder characterized by acid reflux and heartburn—may influence key cardiovascular risk factors, including blood pressure, lipid ...
Repeated exposure to lies online may influence the beliefs of professional content moderators, with consequences for online platforms. Hundreds of thousands of content moderators, typically based in non-Western countries, identify and weed out problematic and false content on social platforms. However, constant exposure to misinformation could convince some content moderators that false claims are true, in what is known as the “illusory truth effect.” Hause Lin and colleagues assessed the extent of this effect among professional content moderators in India and the Philippines and explored whether encouraging an accuracy mindset reduces the effect. ...
COLUMBUS, Ohio – While comics have become a culturally popular and widely studied art form in recent decades, one format remains overlooked: the single-panel comic.
Comics like “The Family Circus,” “Ziggy” and “Little Lulu” are often seen as simplistic and not worthy of critical attention, argues Michelle Ann Abate, author of the new book Singular Sensations: A Cultural History of One-Panel Comics in the United States.
“There tends to be a belief there ...
PULLMAN, Wash. — Unlike kids in the United States, hunter-gatherer children in the Congo Basin have often learned how to hunt, identify edible plants and care for babies by the tender age of six or seven.
This rapid learning is facilitated by a unique social environment where cultural knowledge is passed down not just from parents but from the broader community, according to a new Washington State University-led study in the Proceedings of the National Academy of Sciences.
The research helps explain how many cultural traits have been preserved for thousands of years among ...
US Latinos who rely on social media in Spanish for their news are more vulnerable to political misinformation than those who use English-language social media, according to a study. Misinformation swirls on social media in every language, but social media companies struggle to combat disinformation circulating in Spanish on major social media platforms. In a study done in partnership with Jonathan Nagler of NYU’s Center for Social Media and Politics as part of the group’s Bilingual Election Monitor project, Marisa Abrajano and colleagues surveyed over 1,000 English-dominant, bilingual, and Spanish-dominant Latino users of ...
