Researchers to Test Acceptance of Autonomous AI in Facial Surgery Planning

Researchers for a forthcoming study will examine how increasing levels of artificial intelligence (AI) autonomy in facial reconstructive surgery planning affect public perceptions of acceptability, trust, risk, and responsibility.

Publishing the study protocol in Healthcare, they describe a cross-sectional, within-subjects vignette experiment among Romanian adults. Investigators plan to recruit 300 participants with valid responses (minimum analyzable sample of 140). A 50-participant pilot will precede preregistration and full data collection.

Testing Acceptance of AI in Facial Reconstructive Surgery

Each participant will evaluate four hypothetical facial reconstructive surgery scenarios representing progressively greater AI involvement: imaging assistance, decision support, outcome simulation, and autonomous planning. Participants will be randomly assigned to one of four presentation sequences using a Williams balanced Latin-square design intended to account for serial position and first-order carry-over effects.

After each scenario, participants will rate acceptability, trust, and perceived risk using three 7-point Likert items for each measure. They will also assign responsibility among the surgeon, hospital, AI developer, and unavoidable medical uncertainty. Additional measures will assess perceived AI control, technical sophistication, and scenario difficulty.

The primary analysis will use mixed repeated-measures analyses of variance, including linear, quadratic, and cubic trend contrasts, along with Holm-corrected pairwise comparisons. Investigators also plan a position-adjusted linear mixed model. Secondary analyses will examine shifts in responsibility attribution, AI familiarity, and overall implementation preferences.

Source

Dina V, et al. Healthcare. 2026. Doi:10.3390/healthcare14183030