International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 184 - Number 10 |
Year of Publication: 2022 |
Authors: João Vagner Pereira Da Silva, Sandra Kalil Bussadori, Maria Lucia Z. Varellis, Alessandro M. Deana |
10.5120/ijca2022922008 |
João Vagner Pereira Da Silva, Sandra Kalil Bussadori, Maria Lucia Z. Varellis, Alessandro M. Deana . Classification of Tooth Decay by Computer-Aided Diagnostics using Laser Speckle Image Analysis. International Journal of Computer Applications. 184, 10 ( Apr 2022), 12-17. DOI=10.5120/ijca2022922008
Dental caries is one of the most prevalent diseases in the world, affecting almost the entirety (100%) of the population, generating a great concern in oral health. Its early detection - before the need for invasive restoration procedures -contributes to the general well-being of the population. This work demonstrates the detection and classification of early carious lesions with an ICDAS (International Caries Detection and Assessment System) lower than 1. For this work, we used 45 samples of bovine incisors that were cut, polished, and inserted in a PVC holder. Each sample had half of its surface covered and the other half subjected to an acid etching process producing a carious-like lesion. The samples were illuminated with laser light and imaged. After a pre-processing all images were classified according to the severity of the lesion and a neural network was trained to detect and quantify the lesions. With this approach we obtained over 83% accuracy for the detection of the lesion and over 63% of accuracy for the classification of the lesions that were considered an ICDAS lower than 1 by atrained dentist.