[基于细胞神经网络的图像分割法对牙菌斑的分析]
[Analysis of dental plaque by using cellular neural network-based image segmentation].
作者信息
Luan Qing-xian, Li Xiao, Kang Jia-yin, Liu Jin-zhu, Min Le-quan
机构信息
School of Information Engineering and School of Applied Scinces, University of Science and Technology, Beijing 100083, China.
出版信息
Zhonghua Kou Qiang Yi Xue Za Zhi. 2007 Dec;42(12):720-2.
OBJECTIVE
To establish and evaluate a new method for measurement of dental plaque by using cellular neural network-based image segmentation.
METHODS
A total of 195 subjects were selected from community population. After dental plaque staining, oral digital picture of anterior teeth area was taken by an Olympus digital camera (C-7070 Wide Zoom). At the same time, the Turesky dental plaque indices of anterior teeth were evaluated. The image analysis was conducted by cellular neural network-based image segmentation.
RESULTS
The image cutting errors between two operators were very small. The Kappa value is 0.935. Pearson's correlation coefficient is 0.988 (P < 0.001). There was high correlative consistency between traditional dental plaque index and plaque percentage obtained by using image analysis. Pearson's correlation coefficient was 0.853 (P < 0.001).
CONCLUSIONS
Cellular neural network-based image segmentation is a new method feasible for evaluating dental plaque.
目的
建立并评估一种基于细胞神经网络图像分割的牙菌斑测量新方法。
方法
从社区人群中选取195名受试者。牙菌斑染色后，用奥林巴斯数码相机（C - 7070 Wide Zoom）拍摄前牙区口腔数码照片。同时，评估前牙的Turesky牙菌斑指数。采用基于细胞神经网络的图像分割进行图像分析。
结果
两名操作者之间的图像切割误差非常小。Kappa值为0.935。Pearson相关系数为0.988（P < 0.001）。传统牙菌斑指数与图像分析获得的菌斑百分比之间存在高度相关一致性。Pearson相关系数为0.853（P < 0.001）。
结论
基于细胞神经网络的图像分割是一种评估牙菌斑的可行新方法。
Zotero 插件