Rashid Umer, Javid Aiman, Khan Abdur Rehman, Liu Leo, Ahmed Adeel, Khalid Osman, Saleem Khalid, Meraj Shaista, Iqbal Uzair, Nawaz Raheel
Department of Computer Science, Quaid-e-Azam University, Islamabad, Pakistan.
School of Business and Law, The Manchester Metropolitan University, Manchester, United Kingdom.
PeerJ Comput Sci. 2022 Feb 18;8:e888. doi: 10.7717/peerj-cs.888. eCollection 2022.
Nearly 3.5 billion humans have oral health issues, including dental caries, which requires dentist-patient exposure in oral examinations. The automated approaches identify and locate carious regions from dental images by localizing and processing either colored photographs or X-ray images taken specialized dental photography cameras. The dentists' interpretation of carious regions is difficult since the detected regions are masked using solid coloring and limited to a particular dental image type. The software-based automated tools to localize caries from dental images taken ordinary cameras requires further investigation. This research provided a mixed dataset of dental photographic (colored or X-ray) images, instantiated a deep learning approach to enhance the existing dental image carious regions' localization procedure, and implemented a full-fledged tool to present carious regions simple dental images automatically. The instantiation mainly exploits the mixed dataset of dental images (colored photographs or X-rays) collected from multiple sources and pre-trained hybrid Mask RCNN to localize dental carious regions. The evaluations performed by the dentists showed that the correctness of annotated datasets is up to 96%, and the accuracy of the proposed system is between 78% and 92%. Moreover, the system achieved the overall satisfaction level of dentists above 80%.
近35亿人存在口腔健康问题，包括龋齿，而这需要在口腔检查中医生与患者接触。自动化方法通过定位和处理由专业牙科摄影相机拍摄的彩色照片或X射线图像，从牙科图像中识别和定位龋损区域。由于检测到的区域使用纯色掩盖且仅限于特定的牙科图像类型，医生对龋损区域的解读存在困难。基于软件的从普通相机拍摄的牙科图像中定位龋齿的自动化工具仍需进一步研究。本研究提供了一个牙科摄影（彩色或X射线）图像的混合数据集，实例化了一种深度学习方法以增强现有的牙科图像龋损区域定位程序，并实现了一个成熟的工具来自动在简单牙科图像中呈现龋损区域。该实例化主要利用从多个来源收集的牙科图像（彩色照片或X射线）的混合数据集以及预训练的混合掩码区域卷积神经网络（Mask RCNN）来定位牙科龋损区域。医生进行的评估表明，注释数据集的正确率高达96%，所提出系统的准确率在78%至92%之间。此外，该系统获得了超过80%的医生总体满意度。
