AI-Driven Assessment of Lip Volume Improvement Using Hyaluronic Acid Fillers: A Comprehensive Analysis.

Background: The lips are a popular area for filler treatments across all age groups. Existing evaluations of lip volume enhancement often rely on subjective, visual assessments, and fail to account for the overall facial balance and esthetic changes. To address this, this study uses various AI models for a comprehensive analysis to better evaluate the effects of lip enhancement and provide insights for patient-specific treatment plans.

Methods: This study involved 31 participants (29 females, 2 males), all of whom received lip HA filler injections. Pre- and post-treatment facial images were captured in a relaxed state, and these images were processed using AI-based tools to record and analyze facial visual analysis, perceived age, and emotional expressions displayed on the face.

Results: Post-treatment, the facial visualization analysis revealed that as lip volume increased, the lip line became more defined and the natural curve of the lips became more balanced. The MiVOLO model demonstrated high age estimation accuracy (MAE: 5.15; ICC:0.866; Pearson's r: 0.875), with a mean apparent age reduction of - 1.63 ± 2.51 years (95% CI = -2.78 to - 0.49, Cohen's d = -0.65), emotion analysis revealed a significant increase in neutral expression (72.97-83.44%, 95% CI = 3.38 to 17.56, Cohen's d = 0.58) and reductions in anger (10.25-5.84%, 95% CI = -8.53 to - 2.28, Cohen's d = -0.42).

Conclusions: This study provides significant insights by showing that lip volume improvement affects not only external changes but also emotional expression, age perception, and social interaction. This approach presents a new solution in the field of cosmetic surgery and will contribute to improving patient satisfaction through personalized treatments and multifaceted effect analysis.

Level of evidence iv: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .