Aplicação De Uma Rede Neural Artificial De Base Radial Para Suporte Ao Diagnóstico De Cáries Dentárias Através De Radiografias Applying an Artificial Radial Basis Network for Supporting Radiographic Diagnosis of Dental Caries

Resumo: Este estudo utiliza um modelo de inteligência artificial, uma rede neural com base radial (RBF), para dar suporte a diagnósticos radiográficos de cáries dentárias. Cento e sessenta imagens radiografias de faces proximais de dentes humanos extraídos foram avaliadas no que se refere à presença ou ausência de cárie por vinte e cinco examinadores. As condições reais de cada dente, denominadas gold standards, foram analisadas através de microscopia ótica. Estas gold standards foram empregadas para treinar a rede neural para diagnosticar cáries com base nos exames radiográficos. A fim de avaliar a capacidade da rede para a generalização, verificando o seu desempenho em relação a diagnóstico de novos casos, os dados foram divididos em dois subgrupos: subgrupo de treinamento e subgrupo de teste. Curvas ROC (Receiver Operating Characteristics) permitiram a Abstract: This study uses an artificial intelligent model, a Radial Basis Function neural network (RBF), to support radiography diagnosis of dental caries. One hundred and sixty radiography images of proximal faces of extracted human teeth were analyzed by twenty-five examiners, which diagnosed the presence or absence of dental caries. The same teeth were then subjected to optical microscope analysis, which allowed the verification of their actual conditions. Such information was named as gold standards, and served to train a neural network to diagnose caries by means of radiography images. In order to verify the network's ability to diagnose new cases, data were organized in two subgroups: training subgroup and test subgroup. ROC (Receiver Operating Characteristics) curves allowed the comparison between diagnosis efficacy with or without the use of neural network, showing that the adopted artificial intelligent model significantly improved diagnosis qualities.