AI-Based Facial Analysis vs Self-Report: A Pilot Study Using Insights From Pre-Procedural Psychological Tests

BACKGROUND Aesthetic procedures aligning with societal beauty standards are gaining popularity, often driven by unrealistic beauty ideals. However, these treatments frequently disappoint individuals with low self-esteem and distorted self-images. AI tools offer potential for a more objective assessment, generating pre-treatment simulations. This study explores the relationship between psychological profiles and AI evaluations. MATERIAL AND METHODS Forty-seven participants underwent psychological tests (TIPI-PL, SWLS, Rosenberg's Self-Esteem Scale, and Self-Perception Scale). An AI neural network analyzed facial images across 7 traits: attractiveness, competence, restfulness, joy, dominance, health, and trustworthiness. The AI application dedicated to tablets that we used is available online. After taking a photo of the face, the application creates a heatmap with facial features pictured on the scale. Participants also self-assessed these traits using the SPS scale. Pearson's r and the Mann-Whitney U tests were used to examine correlations and compare AI and self-assessment results. RESULTS The AI-based analysis assessed female faces as more attractive (P<0.001), more rested (P=0.004), happier (P<0.001) and more trustworthy (P<0.001). There was a significant association trending towards statistical significance between 'Competent' and 'Rested' AI results with TIPI Extraversion (P=0.045). None of the other results of the analysis of psychological status, personality types, self-esteem, and well-being reached statistical significance. CONCLUSIONS AI applications present promising opportunities for enhancing medical practices. In this study we observed some limitations, especially concerning the intricate interplay between physical appearance, psychological factors, and AI facial evaluation. The SPS scale was significantly correlated with other psychological tests, although AI ratings were generally lower than self-assessments. Therefore, AI should not be the only tool used for planning aesthetic procedures.