Quantitative Light Fluorescence (QLF) and Polarized White Light (PWL) assessments of dental fluorosis in an epidemiological setting
- Equal contributors
1 Colgate Palmolive Dental Health Unit, School of Dentistry, University of Manchester, Lloyd Street North, Manchester Science Park, England, M15 6SH, United kingdom
2 Oral Health Unit, School of Dentistry, University of Manchester, Lloyd Street North, Manchester Science Park, England, M15 6SH, United kingdom
3 CDC, National Center for Health Statistics, Metro IV Building, 3311 Toledo Road, Hyattsville, MD, 20782, USA
4 NIH, National Institute of Dental and Craniofacial research, 31 Center Drive Ste 5B55, Bethesda, MD, 20892-2190, USA
5 Indiana University, School of Dentistry, Oral Health Research Institute, 415 Lansing St, Indianapolis, IN, 46202-2876, USA
6 Faculty of Dentistry, Chiang Mai University, T. Suthep, A. Muang, Chiang Mai, 50200, Thailand
BMC Public Health 2012, 12:366 doi:10.1186/1471-2458-12-366Published: 20 May 2012
To determine if a novel dual camera imaging system employing both polarized white light (PWL) and quantitative light induced fluorescence imaging (QLF) is appropriate for measuring enamel fluorosis in an epidemiological setting. The use of remote and objective scoring systems is of importance in fluorosis assessments due to the potential risk of examiner bias using clinical methods.
Subjects were recruited from a panel previously characterized for fluorosis and caries to ensure a range of fluorosis presentation. A total of 164 children, aged 11 years (±1.3) participated following consent. Each child was examined using the novel imaging system, a traditional digital SLR camera, and clinically using the Dean’s and Thylstrup and Fejerskov (TF) Indices on the upper central and lateral incisors. Polarized white light and SLR images were scored for both Dean’s and TF indices by raters and fluorescence images were automatically scored using software.
Data from 164 children were available with a good distribution of fluorosis severity. The automated software analysis of QLF images demonstrated significant correlations with the clinical examinations for both Dean’s and TF index. Agreement (measured by weighted Kappa’s) between examiners scoring clinically, from polarized photographs and from SLR images ranged from 0.56 to 0.92.
The study suggests that the use of a digital imaging system to capture images for either automated software analysis, or remote assessment by raters is suitable for epidemiological work. The use of recorded images enables study archiving, assessment by multiple examiners, remote assessment and objectivity due to the blinding of subject status.