文献速递美容整形

AI面部真实性定位模型检测隆鼻术后网络图片篡改

Detecting Manipulated Online Rhinoplasty Images Using an Artificial Intelligence Facial Authenticity Localization (FAL) Model.
PubMed 文献 · Plast Reconstr Surg · 2026-09-21

本研究首次将面部真实性定位(FAL)模型大规模应用于美容外科媒体,分析RealSelf.com公开画廊连续600张隆鼻术后照片。结果显示19.5%的照片存在鼻部区域篡改,验证集假阳性率1.5%,对Facetune几何扭曲的灵敏度为100%。提示平台隆鼻照片约五分之一存在可疑编辑。

要点

  • RealSelf.com隆鼻术后照片中19.5%存在鼻部区域篡改。
  • FAL模型验证假阳性率1.5%,对Facetune几何扭曲灵敏度100%。
  • 建议将自动真实性检查整合入临床摄影与平台流程。

核心结论

  • AI几何篡改检测发现主要平台上约五分之一的公开隆鼻照片存在可疑编辑。
  • 验证显示对Facetune生成的编辑具有高特异度和灵敏度。

研究设计与样本

  • 研究分析RealSelf.com公开画廊2025年10月连续600张隆鼻术后照片,未额外筛选。
  • FAL模型在macOS本地通过Anaconda-Python环境运行,使用预训练全局与局部权重。
  • 特异性在200张 presumed-unedited 临床RAW导出照片上估计,灵敏度在50张故意扭曲的临床照片上评估。

主要发现

  • RealSelf.com篡改患病率为19.5%(117/600;95% CI,16.3-22.7%)。
  • 验证假阳性率为1.5%(3/200;95% CI,0.51-4.32%)。
  • 对Facetune生成几何扭曲的灵敏度为100%(50/50;95% CI,93.0-100%)。

关键数值与推荐

  • 阳性定义为鼻部区域热图激活。
  • 篡改患病率19.5%(117/600;95% CI,16.3-22.7%)。
  • 假阳性率1.5%(3/200;95% CI,0.51-4.32%)。
  • 灵敏度100%(50/50;95% CI,93.0-100%)。

临床怎么用

  • 可将自动真实性检查整合入临床摄影和平台工作流程以提高透明度。
  • 临床中应警惕患者基于网络图片的期望可能受篡改影响。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Publicly available before-and-after photos influence patient expectations and perceptions of surgical success. Online platforms host thousands of images, yet the authenticity of these representations remain unverified in an age where digitally altered photographs are increasingly common in aesthetic surgery. The Facial Authenticity Localization (FAL) model can detect subtle pixel-level irregularities indicative of digital tampering. This study represents the first large-scale application of an AI-based authenticity detector to aesthetic-surgery media. METHODS: The first 600 consecutive postoperative rhinoplasty photographs in RealSelf.com's public gallery (October 2025) were analyzed without additional filtering or selection. FAL was implemented locally on macOS via an Anaconda-Python environment with pretrained global and local weights. Heatmap outputs visualized probable regions of digital manipulation. Positives were defined as nasalregion heatmap activations. Specificity was estimated on 200 presumed-unedited clinical photographs exported from RAW with minimal processing; sensitivity was assessed on 50 clinical, intentionally warped photographs. RESULTS: Manipulation prevalence on RealSelf.com was 19.5% (117/600; 95% CI, 16.3-22.7%). The validation falsepositive rate was 1.5% (3/200; 95% CI, 0.51-4.32%). Sensitivity for detecting Facetune-generated geometric warps was 100% (50/50; 95% CI, 93.0-100%). CONCLUSIONS: AIbased geometric manipulation detection identifies suspicious edits in roughly one in five public rhinoplasty photos on a major platform. Validation suggests high specificity and sensitivity for Facetune-generated edits. Integration of automated authenticity checks into clinical photography and platform workflows may improve transparency.

来源:Plast Reconstr Surg,查看原文。版权归原作者所有。

引用

  1. Olabiran A, Kim R, Rohrich RJ, et al. Detecting Manipulated Online Rhinoplasty Images Using an Artificial Intelligence Facial Authenticity Localization (FAL) Model. Plast Reconstr Surg. 2026. PMID: 42765934. doi:10.1097/PRS.0000000000013464

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。