Division of Restorative Dental Sciences, Faculty of Dentistry, The University of Hong Kong, Sai Ying Pun, Hong Kong SAR, PR China.
Department of Radiology, Northwestern University, Chicago, United States.
J Dent. 2023 Nov;138:104721. doi: 10.1016/j.jdent.2023.104721. Epub 2023 Sep 22.
To determine the agreement of removable dental prosthesis cleanliness as assessed by a semi-automated planimetric method between images captured by a digital single-lens reflex camera (DSLR) and smartphone.
A total of 97 participants with removable prostheses were recruited for the quantification of the prosthesis plaque area coverage. The colour images of stained prosthesis plaque were obtained using both a DSLR camera and a smartphone. The prosthesis plaque area coverage was analysed in two ways: (i) prosthesis cleanliness index (PCI) and (ii) percentage plaque area coverage (PPC). The PPC (continuous data) was converted to the PCI (categorical data) to provide prevalence ordinal scales and the agreements in PCI ratings were determined using weighted Kappa statistics. Agreement of PPC scores was determined through assessing directional, standardised directional, and absolute differences and correlation analyses.
Weighted Kappa values of agreement between PCI categories were excellent (> 0.80) for all comparisons. The mean PPC was 24.79 % as determined by DSLR and 25.37 % as determined by smartphone. There was no statistically significant difference in the means of PPC between the DSLR and smartphone (P = 0.149). The standardised directional difference was 0.15 ('small'). The mean absolute difference was 2.77. The interclass correlation coefficient was 0.98 ('excellent').
This method showed almost perfect agreements and allowed for threshold-based plaque segmentation on the removable prostheses. There was substantial agreement between DSLR and smartphone assessment of prosthesis plaque area coverage as determined by a semi-automated planimetric assessment.
This semi-automated planimetric assessment method has implications for monitoring removable prosthesis hygiene initiatives by offering a valid, reliable, and quantitative method of assessment with potential use in managed care and community settings.
通过半自动平面测量法，评估数码单反相机（DSLR）和智能手机拍摄的图像，确定可摘局部义齿清洁度的一致性。
招募了 97 名佩戴可摘局部义齿的参与者，对义齿菌斑面积进行量化。使用 DSLR 相机和智能手机获取染色义齿菌斑的彩色图像。使用两种方法分析义齿菌斑面积：（i）义齿清洁指数（PCI）和（ii）菌斑面积百分比（PPC）。PPC（连续数据）转换为 PCI（分类数据），提供了患病率等级量表，并使用加权 Kappa 统计确定 PCI 评级的一致性。通过评估方向、标准化方向和绝对差异以及相关分析来确定 PPC 评分的一致性。
所有比较的 PCI 分类之间的加权 Kappa 值均为优秀（>0.80）。通过 DSLR 确定的平均 PPC 为 24.79%，通过智能手机确定的平均 PPC 为 25.37%。两种设备之间的 PPC 均值无统计学差异（P=0.149）。标准化方向差异为 0.15（“小”）。平均绝对差异为 2.77。组内相关系数为 0.98（“优秀”）。
该方法显示出几乎完美的一致性，并允许在可摘局部义齿上进行基于阈值的菌斑分割。半自动平面测量评估的 DSLR 和智能手机评估可摘局部义齿菌斑面积的一致性较高。
这种半自动平面测量评估方法具有监测可摘局部义齿卫生措施的意义，提供了一种有效的、可靠的、定量的评估方法，具有在管理式医疗和社区环境中应用的潜力。
