Abstract | OBJECTIVES: Most dental plaque indices have a subjective component due to investigator interpretation. To assess the accumulation of plaque on teeth, a fully automated method was developed based on digital imaging of methylene blue (1% (w)/(v)) disclosed plaque. METHODS: Intra-oral colour images of disclosed plaque from 25 individuals were coded into hue, saturation and intensity colour space. Unequivocal areas of clean teeth, plaque and clean gingiva were sampled to determine the typical values of hue, saturation and intensity of light. RESULTS: Discriminant analysis showed that the success rate for classification of the pixels as plaque and non-plaque was 87.3% based on saturation and intensity, and 98.7% for a classifier based on hue and intensity. Fisher's function coefficients were used to create classification boundaries in the colour space which would allow to automatically classify new images. CONCLUSIONS: This method has potential for automated and quantitative measurement screening of dental plaque that may be used to assess the efficacy of oral hygiene products and procedures.
|
Authors | Kevin Carter, Gabriel Landini, A Damien Walmsley |
Journal | Journal of dentistry
(J Dent)
Vol. 32
Issue 8
Pg. 623-8
(Nov 2004)
ISSN: 0300-5712 [Print] England |
PMID | 15476956
(Publication Type: Comparative Study, Evaluation Study, Journal Article, Research Support, Non-U.S. Gov't)
|
Chemical References |
|
Topics |
- Adult
- Dental Plaque
(diagnosis)
- Dental Plaque Index
- Diagnosis, Computer-Assisted
(methods)
- Female
- Humans
- Image Processing, Computer-Assisted
(methods)
- Incisor
- Male
- Methylene Blue
- Middle Aged
- Observer Variation
- Reproducibility of Results
- Sensitivity and Specificity
|