建立具有不同鼻形态的三维面部模板数据集及其在制作鼻缺损修复体中的应用
本研究从北京大学口腔医院三维面部数据库中筛选115名男性和115名女性成人数据,构建了12个不同鼻形态的三维面部模板数据集。基于PA-NICP算法,使用该数据集构建的鼻修复体形态与原始鼻形态的平均三维偏差为1.18±0.37毫米,显著小于通用模板的1.55±0.51毫米。该模板数据集在构建鼻修复体形态方面显示出临床适用性。
要点
- 成功创建了包含12个不同鼻形态的三维面部模板数据集。
- 使用模板数据集构建的鼻修复体形态三维偏差(1.18±0.37毫米)显著小于通用模板(1.55±0.51毫米)。
- 在男性鞍鼻和驼峰鼻以及女性鞍鼻、狮鼻、蒜头鼻和驼峰鼻中,实验组三维偏差更小。
核心结论
- 与通用三维面部模板相比,三维面部模板数据集在构建鼻修复体形态方面具有临床适用性。
研究设计与样本
- 从北京大学口腔医院三维面部数据库中筛选115名男性和115名女性成人不同鼻形态的三维面部数据,构建男性和女性各6种鼻形态类型的三维面部模板,形成数据集。
- 基于研究组提出的Procrustes分析-非刚性迭代最近点(PA-NICP)算法,使用60个人工构建的鼻缺损数据集(男女各6种鼻形态类型,每种5个样本),比较使用模板数据集(实验组)与通用模板(对照组)构建的鼻修复体形态。
主要发现
- 成功创建了12个三维面部模板的数据集。
- 对于60个鼻缺损数据集,实验组和对照组构建的鼻修复体形态与原始鼻形态的平均三维偏差分别为1.18±0.37毫米和1.55±0.51毫米,两组差异有统计学意义。
- 对于特定鼻形态,实验组一致显示出更小的三维偏差,尤其在男性鞍鼻和驼峰鼻以及女性鞍鼻、狮鼻、蒜头鼻和驼峰鼻中。
临床怎么用
- 本研究建立的具有不同鼻形态的三维面部模板数据集结合非刚性配准算法,可实现鼻修复体形态的高效数字化设计,显示出良好的临床适用性。
局限性
- 原文未报告局限性。
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
OBJECTIVES: This study aimed to construct three-dimensional (3D) face templates with diverse nasal morphologies, establish a 3D face template dataset, and evaluate its application in fabricating nasal defect prostheses. METHODS: The 3D facial data of 115 adult males and 115 adult females with different nasal morphologies were screened from the 3D facial database of Peking University Hospital of Stomatology. These data were used to construct 3D face templates for males and females with 6 different nasal morphology types, forming a comprehensive 3D face template dataset. Based on the Procrustes Analysis (PA)-Non-Rigid Iterative Closest Point (PA-NICP) algorithm proposed by our research group, nasal prosthesis morphologies were constructed with the template dataset and compared against those constructed with the general template. In total, 60 artificially constructed nasal defect datasets, representing 5 samples for each of the 6 types of nasal morphologies in both males and females, were used. The nasal prosthesis morphologies constructed using the template dataset were used as the experimental group, whereas those constructed using the general template were used as the control group. The 3D deviation between the prosthesis morphology and the original nasal morphology of participants was calculated to compare the effectiveness of the experimental and control groups. RESULTS: A dataset of 12 3D face templates was successfully created. For the 60 nasal defect datasets, the average 3D deviations of the constructed nasal prosthesis morphologies in the experimental and control groups were 1.18 ± 0.37 mm and 1.55 ± 0.51 mm, respectively. The differences between the two groups were statistically significant. For specific nasal morphologies, the experimental group consistently showed smaller 3D deviations, particularly in cases of saddle nose and hump nose for males and saddle nose, lion nose, garlic nose, and hump nose for females. CONCLUSIONS: Compared to a general 3D face template, the 3D face template dataset demonstrated clinical applicability in constructing nasal prosthesis morphology. CLINICAL SIGNIFICANCE: This study established a 3D face template dataset with different nasal morphologies. The template dataset combined with a non-rigid registration algorithm can enable efficient digital design of nasal prosthesis morphology, demonstrating good clinical applicability.
引用
- Wen A, Zhang X, Zhu Y, et al. Establishment of a three-dimensional face template dataset with different nasal morphologies and its application in fabricating nasal defect prostheses. J Dent. 2025. PMID: 40403872. doi:10.1016/j.jdent.2025.105834
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。