Share Email Print
cover

Proceedings Paper • new

Laser speckle imaging for lesion detection on tooth
Author(s): Luciano G. Gavinho; João V. P. Silva; João H. Damazio; Ravana A. Sfalcin; Sidnei A. Araujo; Marcelo M. Pinto; Silvia R. G. Olivan; Renato A. Prates; Sandra K. Bussadori; Alessandro M. Deana
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Computer vision technologies for diagnostic imaging applied to oral lesions, specifically, carious lesions of the teeth, are in their early years of development. The relevance of this public problem, dental caries, worries countries around the world, as it affects almost the entire population, at least once in the life of each individual. The present work demonstrates current techniques for obtaining information about lesions on teeth by segmentation laser speckle imagens (LSI). Laser speckle image results from laser light reflection on a rough surface, and it was considered a noise but has important features that carry information about the illuminated surface. Even though these are basic images, only a few works have analyzed it by application of computer vision methods. In this article, we present the latest results of our group, in which Computer vision techniques were adapted to segment laser speckle images for diagnostic purposes. These methods are applied to the segmentation of images between healthy and lesioned regions of the tooth. These methods have proven to be effective in the diagnosis of early-stage lesions, often imperceptible in traditional diagnostic methods in the clinical practice. The first method uses first-order statistical models, segmenting the image by comparing the mean and standard deviation of the intensity of the pixels. The second method is based on the distance of the chi-square (χ2 ) between the histograms of the image, bringing a significant improvement in the precision of the diagnosis, while a third method introduces the use of fractal geometry, exposing, through of the fractal dimension, more precisely the difference between lesioned areas and healthy areas of a tooth compared to other methods of segmentation. So far, we can observe efficiency in the segmentation of the carious regions. A software was developed for the execution and demonstration of the applicability of the models

Paper Details

Date Published: 8 February 2018
PDF: 8 pages
Proc. SPIE 10473, Lasers in Dentistry XXIV, 1047306 (8 February 2018); doi: 10.1117/12.2286408
Show Author Affiliations
Luciano G. Gavinho, UNINOVE (Brazil)
João V. P. Silva, UNINOVE (Brazil)
João H. Damazio, UNINOVE (Brazil)
Ravana A. Sfalcin, UNINOVE (Brazil)
Sidnei A. Araujo, UNINOVE (Brazil)
Marcelo M. Pinto, UNINOVE (Brazil)
Silvia R. G. Olivan, UNINOVE (Brazil)
Renato A. Prates, UNINOVE (Brazil)
Sandra K. Bussadori, UNINOVE (Brazil)
Alessandro M. Deana, UNINOVE (Brazil)


Published in SPIE Proceedings Vol. 10473:
Lasers in Dentistry XXIV
Peter Rechmann; Daniel Fried, Editor(s)

© SPIE. Terms of Use
Back to Top