AutoFFS: Adversarial Deformations for Facial Feminization Surgery Planning
Researchers have developed AutoFFS, a new framework using adversarial deformations to aid in facial feminization surgery planning. This data-driven approach transforms skull morphologies towards a female appearance by targeting sex classifiers. The generated counterfactual skulls offer quantitative guidance for surgeons, with validation through classifier evaluations, new distance metrics (MFD, MKD), and a human perceptual study confirming the desired characteristics. AI
IMPACT Provides a quantitative, data-driven tool to assist surgeons in planning facial feminization surgeries, potentially improving outcomes for transgender and gender diverse patients.