目的：使用图像分析软件对牙菌斑染色图像进行分析，验证该技术的可操作性和实用性，考察该方法的可重复性，并评价不同菌斑染色方式对结果的影响。方法：从北京大学医学部口腔医学专业新入校学生中招募符合纳入标准的志愿者30人，菌斑染色后拍摄标准化口内前牙区数字图像，用Image Pro Plus 7.0软件进行图像分析，同时评价前牙区的Quigley-Hein菌斑指数。结果：图像分析计算所得菌斑染色面积百分比与临床检查记录的相应牙齿菌斑指数有较高的相关一致性，Spearman相关系数为0.776（P<0.01）。两名研究者用软件计算牙齿面积和菌斑面积的同类相关系数分别为0.956和0.930（P<0.01），Bland-Altman分析图显示仅有极个别的点在95％一致性界限以外。不同菌斑染色的图像分析结果显示，牙齿面积在两组间差异无统计学意义，但菌斑面积差异有统计学意义（P<0.01）。结论：该方法较易于操作掌握，所计算的菌斑百分比与传统的菌斑指数有较高的一致性，且重复性好，不同菌斑染色方法对图像分割结果影响不大，宜选用对菌斑敏感的菌斑染色剂进行图像分析。

Objective：To analyze the plaque staining image by using image analysis software, to verify the maneuverability, practicability and repeatability of this technique, and to evaluate the influence of different plaque stains. Methods:In the study, 30 volunteers were enrolled from the new dental students of Peking University Health Science Center in accordance with the inclusion criteria. The digital images of the anterior teeth were acquired after plaque stained according to filming standardization.The image analysis was performed using Image Pro Plus 7.0, and the Quigley-Hein plaque indexes of the anterior teeth were evaluated. Results:The plaque stain area percentage and the corresponding dental plaque index were highly correlated,and the Spearman correlation coefficient was 0.776 (P<0.01). Intraclass correlation coefficients of the tooth area and plaque area which two researchers used the software to calculate were 0.956 and 0.930 (P<0.01).The Bland-Altman analysis chart showed only a few spots outside the 95% consistency boundaries. The different plaque stains image analysis results showed that the diffe-rence of the tooth area measurements was not significant, while the difference of the plaque area measurements significant (P<0.01). Conclusion:This method is easy in operation and control,highly related to the calculated percentage of plaque area and traditional plaque index, and has good reproducibility.The different plaque staining method has little effect on image segmentation results.The sensitive plaque stain for image analysis is suggested.