Evaluation of the aesthetic parameters of artificial intelligence generated digital smile designs

Aim: The aim of this study was to comparatively evaluate how digital smile designs created by artificial intelligence (AI)–based software and dental technicians with different levels of experience are perceived aesthetically by participant groups with various demographic characteristics. Materials and Methods: Two AI-based software programs (3Shape TRIOS Smile Design and Smile.Design) and two dental technicians with different experience levels (20 years and 3 years) were used to create a total of four distinct digital smile designs. These designs were applied to a female and a male volunteer using the mock-up technique and subsequently photographed. A survey was administered to 653 participants via Google Forms. Participants provided demographic information such as age, gender, education level, specialty, and history of esthetic treatment. Each design was rated using a 5-point Likert scale across four main esthetic parameters (Tooth Volume/Size, Tooth Shape/Form, Facial Harmony, and Overall Esthetics). Additionally, participants ranked the four designs from most to least esthetic. Statistical analyses were performed using IBM SPSS software. One-Way ANOVA, Post Hoc Tukey HSD, and Chi-square tests were applied for group comparisons, with a significance level set at p<0.05. Results: Of the participants, 63.9% were female and 36.1% were male, with 71.2% aged between 18 and 25 years. Statistically significant differences were observed among the evaluation groups in terms of esthetic perception (p<0.05). In overall mean scores, the 3Shape AI-assisted design received the highest ratings, while the experienced technician's design received the lowest (p=0.003). In the female model, the less-experienced technician's design scored significantly higher and was perceived as the most esthetic (p<0.001). Conversely, in the male model, the same design received significantly lower scores than all others, being rated the least esthetic (p<0.001). Gender-based rankings also revealed significant differences. In the female model, female participants preferred the 3Shape AI design, whereas male participants favored the design created by the less-experienced technician (p=0.009). For the male model, female participants selected the Smile.Design AI design as the most esthetic, while male participants again preferred the less-experienced technician's design (p=0.004). Conclusion: Esthetic perception is a subjective phenomenon that varies according to the evaluator's demographic characteristics such as gender, age, and education. There is no universal formula for the "ideal" smile design; esthetic success is deeply influenced by contextual factors such as the gender of the face to which the design is applied. Although AI systems demonstrate strong potential in generating initial designs aligned with general esthetic norms, human intuition and experience remain indispensable—particularly in achieving personalization and meeting gender-specific esthetic expectations. In clinical practice, the final esthetic decision should be made by the clinician, taking into account the patient's individual characteristics and preferences. Keywords: Artificial Intelligence, Digital Smile Design, Esthetic Dentistry, Smile Esthetics, Mock-up, Survey Study.