Liu Keyuan, Elbatel Marawan, Chu Guang, Shan Zhiyi, Sum Fung Hou Kumoi Mineaki Howard, Hung Kuo Feng, Zhang Chengfei, Li Xiaomeng, Yang Yanqi
Division of Paediatric Dentistry and Orthodontics, Faculty of Dentistry, The University of Hong Kong, Hong Kong SAR, China.
Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology, Hong Kong Special Administrative Region, Hong Kong SAR, China.
Sci Data. 2025 Jun 14;12(1):1007. doi: 10.1038/s41597-025-05348-3.
Fenestration and dehiscence (FD) pose significant challenges in dental treatments as they adversely affect oral health. Although cone-beam computed tomography (CBCT) provides precise diagnostics, its extensive time requirements and radiation exposure limit its routine use for monitoring. Currently, there is no public dataset that combines intraoral photographs and corresponding CBCT images; this limits the development of deep learning algorithms for the automated detection of FD and other potential diseases. In this paper, we present FDTooth, a dataset that includes both intraoral photographs and CBCT images of 241 patients aged between 9 and 55 years. FDTooth contains 1,800 precise bounding boxes annotated on intraoral photographs, with gold-standard ground truth extracted from CBCT. We developed a baseline model for automated FD detection in intraoral photographs. The developed dataset and model can serve as valuable resources for research on interdisciplinary dental diagnostics, offering clinicians a non-invasive, efficient method for early FD screening without invasive procedures.
牙开窗和牙裂（FD）在牙科治疗中带来了重大挑战，因为它们会对口腔健康产生不利影响。尽管锥形束计算机断层扫描（CBCT）能提供精确的诊断，但它所需时间长且辐射量大，限制了其在监测方面的常规应用。目前，尚无将口腔内照片与相应CBCT图像相结合的公共数据集；这限制了用于自动检测FD及其他潜在疾病的深度学习算法的开发。在本文中，我们展示了FDTooth数据集，该数据集包含了241名年龄在9至55岁之间患者的口腔内照片和CBCT图像。FDTooth在口腔内照片上标注了1800个精确的边界框，并从CBCT中提取了金标准真值。我们开发了一个用于在口腔内照片中自动检测FD的基线模型。所开发的数据集和模型可为跨学科牙科诊断研究提供宝贵资源，为临床医生提供一种无需侵入性操作的非侵入性、高效的早期FD筛查方法。
