Public Perceptions of Artificial Intelligence and Robotics in Medicine

Abstract

Objective: To understand better the public perception and comprehension of medical technology such as artificial intelligence (AI) and robotic surgery. In addition to this, to identify sensitivity to their use to ensure acceptability and quality of counseling. Subjects and Methods: A survey was conducted on a convenience sample of visitors to the MN Minnesota State Fair (n = 264). Participants were randomized to receive one of two similar surveys. In the first, a diagnosis was made by a physician and in the second by an AI application to compare confidence in human and computerbased diagnosis. Results: The median age of participants was 45 (interquartile range 28-59), 58% were female (n = 154) vs 42% male (n = 110), 69% had completed at least a bachelor's degree, 88% were Caucasian (n = 233) vs 12% ethnic minorities (n = 31) and were from 12 states, mostly from the Upper Midwest. Participants had nearly equal trust in AI vs physician diagnoses. However, they were significantly more likely to trust an AI diagnosis of cancer over a doctor's diagnosis when responding to the version of the survey that suggested that an AI could make medical diagnoses (p = 9.32e-06). Though 55% of respondents (n = 145) reported that they were uncomfortable with automated robotic surgery, the majority of the individuals surveyed (88%) mistakenly believed that partially autonomous surgery was already happening. Almost all (94%, n = 249) stated that they would be willing to pay for a review of medical imaging by an AI if available. Conclusion: Most participants express confidence in AI providing medical diagnoses, sometimes even over human physicians. Participants generally express concern with surgical AI, but they mistakenly believe that it is already being performed. As AI applications increase in medical practice, health care providers should be cognizant of the potential amount of misinformation and sensitivity that patients have to how such technology is represented.

Patient and general public attitudes towards clinical artificial intelligence: a mixed methods systematic review

“…27 The other eight studies recruited participants outside of a health-care context: three recruited university students or affiliates, or both, 26,32,35 and five sampled the general population. 16,24,34,36,39 Among the quantitative and mixed methods studies, ten recruited convenience samples of participants, 14,23,28,30,31,33-37 five did anonymous online surveys for which the response rate could not be calculated, 16,26,27,29,32 three recruited all eligible patients, 12,15,38 one did a simple random sampling of mobile telephone users, 39 and one identified all relevant social media posts. 24 Regarding the type of AI being studied, nine (39%) studies assessed a hypothetical AI to be used in a given clinical scenario, eight (35%) assessed AI that was broadly defined, and six (26%) assessed currently available or soon-to-be available AI tools.…”

“…After using these tools in their intended clinical setting, patients expressed a high satisfaction with the EyeGrader system for detecting diabetic retinopathy (92/96 [96%] reported they were satisfied or very satisfied with automated screening, on the day of screening), 28 the FlorenceD2W-T2 automated fully closed-loop insulin delivery prototype (61/62 [98%] reported they were happy to have their glucose blood concentration controlled automatically), 12 and the Isabel symptom checker, which presents possible diagnoses based on symptoms entered by the user online (278/304 [91%] reported they would use it again). 29 Most participants supported the use of or favourably viewed AI that would be used in medicine generally, 24,27,35,37,38 be used for symptom assessment, 30 be used as part of an ophthalmic device, 39 provide a second opinion for an imaging study, 36 monitor potential complications during surgery, 33 simplify medical notes, 23 screen for skin cancer, 15,20,27 remotely monitor chronic conditions, 15 guide physical therapy, 15 and answer emergency telephone calls. 15 However, the studies presented a diverse range of participant views; for instance, a few (approximately 20%) patients were opposed to the use of biomedical devices and AI-based tools in all four presented scenarios (the use of AI to screen for cancer, to remotely monitor chronic conditions to predict exacerbations, to guide physical therapy through smart clothes, and to answer emergency calls through chatbots), 15 and 22% of patients preferred manual over automated screening for diabetic retinopathy at 1 month follow-up, citing trust as the key reason.…”

“…28 Participants were ambivalent toward the use of AI chatbots in health care 32 and AI that would completely replace radiologists. 14 Overall, the participants had negative views on automated antibiotic dosing 34 and were unwilling to undergo autonomous or partly autonomous surgery 33,36 or use AI chatbots for sexual health advice. 31 Many participants viewed currently available AI as premature technology, 14,24,25,32,33 although some mistakenly believed that AI was already being used for surgery.…”

Patient apprehensions about the use of artificial intelligence in healthcare

“…Currently, very little research has been done characterizing patient and other stakeholder perspectives on applications of AI in healthcare. Additionally, the few studies that have assessed patient perspectives have focused on a narrow array of AI tools, which limits their utility as a guide in anticipating patient engagement with other AI applications in healthcare 13,14 . While engaging patients around specific applications of AI is a crucial step in the research and development process, engagement at this level of specificity does not facilitate analysis of broader public perspectives on AI and its application in healthcare, which is much needed for health policy development, innovation priority setting, and implementation design.…”

Human-Robot Collaboration in the Society of the Future: A Survey on the Challenges and the Barriers

“…A poll of visitors of the Minnesota State Fair showed that slightly more than half of respondents (55%) felt uncomfortable about the automated robotic surgery, with the vast majority of respondents mistakenly thinking, that autonomous robotic surgery is already underway. However, the trust of medical AI is much greater: most respondents are willing to trust AI services for medical imaging analysis and providing diagnoses [20].…”