D13. Harnessing Artificial Intelligence to Study Aesthetic Ratios of the Face
PURPOSE: This study explores the use of artificial intelligence (AI) in measuring craniometric ratios for aesthetic assessments and surgical planning in plastic surgery. METHODS: An AI model to automatically identify craniometric landmarks and calculate aesthetic ratios using a combination of 2D to 3D face meshing and landmark detection techniques. The model was trained on existing datasets. Following model development, we analyzed historical portraits, including Botticelli’s Birth of Venus (thought to be Simonetta Vespucci) to validate the model with aesthetic landmarks and published data. RESULTS: The AI model was able to successfully generate 3D representations from 2-D images. The model was able to identify craniometric landmarks and calculate aesthetic ratios, with discrepancies between AI-generated and manually measured ratios being minimal. This model can be employed to explore the significance of historically significant aesthetic ratios and test theories such as the golden ratio, contributing to a deeper understanding of beauty in facial proportions. CONCLUSION: The AI model was able to analyze Simonetta Vespucci’s face with an acceptable degree of accuracy. AI-based craniometric analysis shows great potential for enhancing efficiency and accuracy in aesthetic studies. Future directions include expanding datasets to study aesthetic ratios, developing personalized aesthetic ratio calculators, coding to search for specific ratios, and exploring the subjective versus objective nature of beauty by examining societal and individual perceptions of aesthetic ratios.
Authors
- Alexander H. Chang (ORCID: https://orcid.org/0000-0001-9036-100X)
- Adam Walchak
- Sameer Patel (ORCID: https://orcid.org/0000-0002-3827-3560)
- Gabrielle Knauer
Institutions
- Fox Chase Cancer Center (US)
- Philadelphia University (US)
Publication Details
- Journal
- Plastic & Reconstructive Surgery Global Open
- Published
- 2025-05-01
- DOI
- https://doi.org/10.1097/01.gox.0001114588.50974.92
- Primary Topic
- Digital Imaging in Medicine
- Type
- article
- Field-Weighted Citation Impact
- 0.00