Clinical validation and assessment of a modular fluorescent imaging system and algorithm for rapid detection and quantification of dental plaque

Abstract

Background: Significant numbers of adults and children have untreated plaque due to poor oral hygiene and consequently suffer from associate dental and systemic diseases.

Methods: A handheld device equipped with 405 nm light-emitting diodes was constructed to examine the prevalence of red fluorescence signatures associated with dental plaque. This device was used for in vivo imaging of all four incisors and all four canines of twenty-eight consenting human subjects. The same areas were further imaged under white light illumination with a commercial image-processing based plaque-imaging device, and evaluated by a hygienist and dentist. A custom computer vision algorithm using pixel information was developed to calculate plaque coverage ratios ranging from 0 (no plaque) to 1 (complete plaque coverage) for images captured by both devices.

Results: The algorithm calculated red fluorescence-based plaque coverage ratios ranging from 0.011 to 0.211 for the subjects imaged. Clinical assessment and statistical analyses of associated plaque ratios of the 405 nm device images indicated high sensitivity and specificity in detecting dental plaque by the experimental device compared to the commercial reference device.

Conclusions: The low-cost and open source 405 nm device and the associated computer vision algorithm successfully captured red fluorescence signatures associated with dental plaque and demonstrated comparable performance to a commercially available device. Therefore, a proof of concept validation was provided for the construction and application of a sensitive cost-effective plaque-detecting device. A miniaturized mobile adaptable version of the device was also provided, together with and a step-by-step guide for device assembly and webhost the associated software, to facilitate open-source access to a cost-effective at-home, in-clinic oral care technology.

Trial registration: ClinicalTrials.gov NCT03379337, December 19 2017. Retrospectively registered.

Keywords: Dental plaque; Fluorescence; Oral health; Oral imaging; Porphyrin; Segmentation.

Conflict of interest statement

Ethics approval and consent to participate

Ethics approval was given by the Committee on the Use of Humans as Experimental Subjects at the Massachusetts Institute of Technology, and the protocol indexed as 1,603,518,893. Dr. Edlund is the owner of Hampden Dental Care and provided permission to use this site for data collection. All subjects were verbally informed and provided written details of the study. All subjects provided written consent to participate in the study and for the use of their de-identified images and video for publication.

Consent for publication

All subjects provided written consent to participate in the study and for the use of their de-identified images and video for publication.

Competing interests

All authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures

Fig. 1

The Plaquefinder imaging device and its smartphone attachment variant. a device component view; b assembled view; c interchangeable head component view; d interchangeable head assembled view; e device with interchangeable head attached; f smartphone attachment component view; g smartphone attachment assembled view; h connected and powered smartphone attachment; i Plaquefinder smartphone attachment and original device (wire/electrical connections and joints not shown). Items A through G are CAD renderings

Fig. 2

Image acquisition process during clinical study

Fig. 3

Image post-processing workflow for segmenting dental plaque

Fig. 4

Comparison images captured by reference device and Plaquefinder. Columns from left to right: reference device white light; reference device plaque mode; Plaquefinder white light; Plaquefinder 405 nm light, segmented Plaquefinder 405 nm light images. In the Plaquefinder 405 nm images, teeth have a natural green fluorescence; plaque fluoresces red. Top row, a-e: plaque coverage of 0.2708 (M1). Middle row, F-J: plaque coverage of 0.1891 (M2). Bottom row, K-O: plaque coverage of 0.0213 (M3). Spectroscopic readings from M1 and M2’s red fluorescence are available in an additional document [see Additional file 2]

Fig. 5

Special cases. Top row: a White light image of gingival recession; the dentin root of the tooth has been exposed. b The reference device exaggerated the natural yellow color of the dentin to look like plaque; however, (c) the Plaquefinder showed no visible red fluorescence. Hygienist confirmation of these teeth as clinically plaque-free demonstrates higher specificity in the Plaquefinder. Middle and bottom row: Dental restoration under different imaging modes: d Reference device white light; e Reference device plaque mode; f Plaquefinder white light; g Plaquefinder 405 nm light. The natural green fluorescence of the tooth is contrasted greatly from the restoration in (g). Bacterial activity is visible at the margin as indicated by red fluorescence, a common precursory location for secondary caries