Department of Oral-Maxillofacial Surgery and Orthodontics, The University of Tokyo Hospital, Tokyo, Japan.
School of Engineering Medicine, Beihang University, Beijing, China.
Int J Med Robot. 2024 Jun;20(3):e2651. doi: 10.1002/rcs.2651.
Quantitative evaluation of facial aesthetics is an important but also time-consuming procedure in orthognathic surgery, while existing 2D beauty-scoring models are mainly used for entertainment with less clinical impact.
A deep-learning-based 3D evaluation model DeepBeauty3D was designed and trained using 133 patients' CT images. The customised image preprocessing module extracted the skeleton, soft tissue, and personal physical information from raw DICOM data, and the predicting network module employed 3-input-2-output convolution neural networks (CNN) to receive the aforementioned data and output aesthetic scores automatically.
Experiment results showed that this model predicted the skeleton and soft tissue score with 0.231 ± 0.218 (4.62%) and 0.100 ± 0.344 (2.00%) accuracy in 11.203 ± 2.824 s from raw CT images.
This study provided an end-to-end solution using real clinical data based on 3D CNN to quantitatively evaluate facial aesthetics by considering three anatomical factors simultaneously, showing promising potential in reducing workload and bridging the surgeon-patient aesthetics perspective gap.
面部美学的定量评估是正颌手术中的一个重要但耗时的程序，而现有的二维美学评分模型主要用于娱乐，对临床的影响较小。
设计并训练了一个基于深度学习的 3D 评估模型 DeepBeauty3D，使用了 133 名患者的 CT 图像。定制的图像预处理模块从原始 DICOM 数据中提取骨骼、软组织和个人身体信息，预测网络模块采用 3 输入 2 输出卷积神经网络（CNN）接收上述数据并自动输出美学评分。
实验结果表明，该模型从原始 CT 图像中以 0.231±0.218（4.62%）和 0.100±0.344（2.00%）的准确度预测了骨骼和软组织评分，耗时为 11.203±2.824s。
本研究使用基于 3D CNN 的真实临床数据提供了一个端到端的解决方案，通过同时考虑三个解剖因素来定量评估面部美学，有望减少工作量并弥合外科医生和患者美学观点之间的差距。
Zotero 插件