基于二维面部照片的人工智能自动诊断与分类骨骼性错颌畸形:多模态深度学习方法
本研究回顾性收集1427例患者的2854张面部照片及侧位头影测量片和元数据,开发两阶段深度学习模型FaceDSM-Net(以MobileNetV3-Large为骨干),用于基于二维面部照片识别矢状向和垂直向骨骼性错颌畸形。模型在矢状向分类中AUC为0.96、准确率0.85,垂直向分类中AUC为0.95、准确率0.86,性能与72名不同临床经验评估者相当,具有良好可解释性和中等泛化能力。
要点
- FaceDSM-Net基于二维面部照片联合性别和年龄元数据,可分类矢状向与垂直向骨骼性错颌畸形。
- 模型矢状向分类AUC 0.96、准确率0.85;垂直向分类AUC 0.95、准确率0.86。
- 模型性能与72名不同临床经验评估者相当,可作为临床决策支持工具。
核心结论
- FaceDSM-Net利用多模态数据有效分类矢状向和垂直向骨骼性错颌畸形,具有良好可解释性和中等泛化能力,尤其在正面垂直向任务中表现较好。
研究设计与样本
- 回顾性收集1427例患者的2854张面部照片,以及对应侧位头影测量片和元数据(性别、年龄)。
- 开发两阶段深度学习模型FaceDSM-Net,以MobileNetV3-Large为骨干,整合面部照片与元数据。
- 模型进行内部训练和测试,并通过独立外部测试集评估泛化能力。
- 使用准确率、精确率、AUC、召回率和F1分数评估性能,采用Grad-CAM进行可解释性分析。
- 将模型性能与72名不同临床经验评估者在同一数据集上的表现进行比较。
主要发现
- FaceDSM-Net在矢状向分类中AUC为0.96、准确率为0.85。
- FaceDSM-Net在垂直向分类中AUC为0.95、准确率为0.86。
- 纳入性别和年龄元数据提高了模型的稳健性。
- 热力图显示FaceDSM-Net聚焦于与骨骼性错颌畸形相关的关键面部结构。
- 模型与人类评估者相比表现出可比的可靠性。
临床怎么用
- 该模型可减少资源和成本负担,同时保持诊断性能。
- 支持早期筛查、个体和家庭层面诊断以及纵向随访。
- 与专家评估者性能相当,可作为可靠的临床决策支持工具,尤其适用于非放射学或资源有限环境。
局限性
- 模型泛化能力为中等,尤其在非正面垂直向任务中可能受限。
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
OBJECTIVE: This study aims to establish and validate a two-stage deep learning framework that enables reliable and cost-effective identification of sagittal and vertical skeletal malocclusions based on 2D facial photographs. METHODS: 2854 photographs were retrospectively collected from 1427 patients, along with corresponding lateral cephalograms and metadata (sex and age). A two-stage deep learning model (FaceDSM-Net) was developed using MobileNetV3-Large as the backbone, integrating facial photographs with metadata. The model was trained and tested internally, with performance evaluated by accuracy, precision, AUC, recall, and F1-score. Grad-CAM was used for interpretability. The generalizability was assessed by an independent external test set, and the performance was further compared with that of 72 raters with varying clinical experience on the same dataset. RESULTS: FaceDSM-Net demonstrated excellent performance in sagittal (AUC 0.96, accuracy 0.85) and vertical (AUC 0.95, accuracy 0.86) classifications. Its robustness was improved by incorporating sex and age. Heatmaps showed FaceDSM-Net focused on key facial structures associated with skeletal malocclusion. The model demonstrated comparable reliability compared to human raters. CONCLUSION: FaceDSM-Net effectively classified sagittal and vertical skeletal malocclusions using multimodal data, demonstrating good interpretability and moderate generalizability, particularly in the frontal-vertical task. It holds potential for clinical application in skeletal malocclusion prediction. CLINICAL SIGNIFICANCE: This model significantly reduces resource and cost burdens while maintaining diagnostic performance. It supports early screening, individual and family-level diagnosis, and longitudinal follow-up. Its comparable performance over expert raters highlights its potential as a reliable clinical decision support tool, particularly in non-radiographic or resource-limited settings.
引用
- Yang H, Bu WQ, Shu YX, et al. Automated diagnosis and classification of skeletal malocclusion using artificial intelligence on 2D facial photographs: A multi-modal deep learning approach. J Dent. 2025. PMID: 41270953. doi:10.1016/j.jdent.2025.106257
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。