Attitudes of the Surgical Team Toward Artificial Intelligence in Neurosurgery: International 2-Stage Cross-Sectional Survey

Abstract

Background Artificial intelligence (AI) has the potential to disrupt how we diagnose and treat patients. Previous work by our group has demonstrated that the majority of patients and their relatives feel comfortable with the application of AI to augment surgical care. The aim of this study was to similarly evaluate the attitudes of surgeons and the wider surgical team toward the role of AI in neurosurgery. Methods In a 2-stage cross sectional survey, an initial open-question qualitative survey was created to determine the perspective of the surgical team on AI in neurosurgery including surgeons, anesthetists, nurses, and operating room practitioners. Thematic analysis was performed to develop a second-stage quantitative survey that was distributed via social media. We assessed the extent to which they agreed and were comfortable with real-world AI implementation using a 5-point Likert scale. Results In the first-stage survey, 33 participants responded. Six main themes were identified: imaging interpretation and preoperative diagnosis, coordination of the surgical team, operative planning, real-time alert of hazards and complications, autonomous surgery, and postoperative management and follow-up. In the second stage, 100 participants responded. Responders somewhat agreed or strongly agreed about AI being used for imaging interpretation (62%), operative planning (82%), coordination of the surgical team (70%), real-time alert of hazards and complications (85%), and autonomous surgery (66%). The role of AI within postoperative management and follow-up was less agreeable (49%). Conclusions This survey highlights that the majority of surgeons and the wider surgical team both agree and are comfortable with the application of AI within neurosurgery.

Neurosurgery and artificial intelligence

“…As such, telemedicine can be used for follow-up visits which would have required in-person clinic visits, remote diagnosis of postsurgery complications, as well as continued routine monitoring [89] . In comparison, the use of AI in postoperative care and follow-up of neurosurgery is less agreeable among surgeons and surgical teams compared to other usages such as pre-operative planning and image interpretation [90] .…”

“…Recent qualitative studies have shown that while patients and their relatives welcome the use of AI in neurosurgery and find it acceptable, they were reluctant to be treated by fully autonomous neurosurgeries and would like to have neurosurgeons to ultimately remain in control [104] . Similarly, most neurosurgeons welcome the use of AI in neurosurgery [90] , so there is a consensus among neurosurgeons and patients on the use of AI in neurosurgery. Multiple studies have shown that the combination of AI with clinicians improved clinical decision making compared to clinicians or AI alone [50] , [70] , [105] , [106] .…”

Artificial Intelligence in Brain Tumour Surgery—An Emerging Paradigm

“…The widespread introduction of AI in brain tumour surgery will require the addition of AI to the neurosurgical and oncological teaching curriculum. At present, neurosurgical teaching curricula has little focus on AI as a teaching subject, with research demonstrating that one third have no prior knowledge of the subject [ 203 ]. As clinicians, we must be expected to understand the mechanisms of the technology we operate, such that black-box systems are not perpetuated.…”

“…Concerns regarding clinician displacement and replacement by AI robotics —the “human-vs.-machine” paradigm—are unlikely to bear truthful in the near future, and indeed, research has identified that neurosurgeons are generally comfortable with greater integration of AI [ 8 , 203 ]. Humans and ML models are likely to work in tandem with one another, rather than be fully replaced.…”

Survey to healthcare professionals on the practicality of AI services for healthcare

“…The survey was created using Google Forms. A link to the survey was sent to healthcare professionals by informing healthcare professionals in our own network in Finland via Facebook and using SurveyCircle scientific survey service 10 . No individual requests to complete the survey were sent.…”