Most patients trust AI to interpret their imaging, but certain demographic factors shape these opinions
Patients are cautiously optimistic about artificial intelligence applications playing a role in their imaging interpretations, according to new survey data.
Research in RSNA journal Radiology: Imaging Cancer details the opinions of a diverse population of patients pertaining to the use of AI in radiology. The large sample size offers new insight into how different demographics, medical histories and educational backgrounds influence an individual’s trust of AI—information that is of great value for organizations hoping to implement AI into clinical practice.
“Patient perspectives are crucial because successful AI implementation in medical imaging depends on trust and acceptance from those we aim to serve,” noted study author Basak E. Dogan, MD, a clinical professor of radiology and director of breast imaging research at the University of Texas Southwestern Medical Center in Dallas. “If patients are hesitant or skeptical about AI’s role in their care, this could impact screening adherence and, consequently, overall healthcare outcomes.”
To gauge patients’ perceptions of AI, Dogan and colleagues distributed a voluntary 29-question survey to every patient who completed a breast cancer screening exam at their facility during seven months in 2023. Participants who completed the survey also shared information on their age, race and ethnicity, education, income level, and history of breast cancer and biopsy.
More than 500 patients completed the survey. Of those, 71% supported the use of AI as a second reader, while just 4.4% were comfortable with algorithms flying solo.
The opinions varied based on knowledge of AI, education level, and race and ethnicity. Individuals with college degrees or greater self-reported AI knowledge were twice as likely to support the use of AI, while Hispanic and non-Hispanic Black respondents were more skeptical due to concerns about AI bias and data privacy. Patients with a personal or family history also were more hesitant to trust AI reads and were more likely to request additional reviews of their imaging.
“Our study shows that trust in AI is highly individualized, influenced by factors such as prior medical experiences, education and racial background,” Dogan said, adding that it is important to consider these perspectives when developing implementation strategies to ensure AI will “improve and not hinder patient care.”
The study abstract is available here.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She began covering the medical imaging industry for Innovate Healthcare in 2021.