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

Area of Science:

• Medical Aesthetics
• Artificial Intelligence
• Psychology

Background:

• Aesthetic procedures are popular but can disappoint individuals with low self-esteem due to unrealistic beauty standards.
• Artificial intelligence (AI) offers objective assessment potential for pre-treatment simulations in aesthetic medicine.
• Understanding the interplay between psychological profiles and AI facial evaluations is crucial.

Purpose of the Study:

• To explore the relationship between psychological profiles and AI-driven facial evaluations.
• To compare AI assessments of facial traits with self-assessments.

Main Methods:

• 47 participants completed psychological tests (TIPI-PL, SWLS, Rosenberg's Self-Esteem Scale, Self-Perception Scale).
• An AI neural network analyzed facial images for 7 traits: attractiveness, competence, restfulness, joy, dominance, health, and trustworthiness.
• Correlations and comparisons between AI and self-assessments were analyzed using Pearson's r and Mann-Whitney U tests.

Main Results:

• AI rated female faces as significantly more attractive, rested, happy, and trustworthy.
• A trend suggested an association between 'Competent' and 'Rested' AI results and TIPI Extraversion.
• No other significant correlations were found between psychological status and AI facial evaluations.

Conclusions:

• AI holds promise for medical applications, but limitations exist in evaluating the complex relationship between appearance and psychological factors.
• While the Self-Perception Scale correlated with psychological tests, AI ratings were generally lower than self-assessments.
• AI should be used cautiously and not as the sole tool for planning aesthetic procedures.