Volume 05,Issue 02

A New Method for Dental Caries Diagnosis Using Convolutional Neural Networks and Bees Algorithm

Authors

Abdolhamid Anaei, Abdol Aziz Kalteh

Abstract
Over the past few years, dental radiography has played an important role in clinical diagnosis, treatment and surgery. Accordingly, extensive efforts have been done on improving computerized dental X-ray image analysis systems for clinical usages. This paper proposes a new method based on convolutional neural networks (CNNs) and bee's algorithm (BA) for detection and diagnosis of dental caries on X-ray image. The proposed method includes two main modules: classifier module and optimization module. In the classifier module, the CNN is used as the main classifier. One of the advantages of CNN is feature representation that is learned automatically from the training data, which is a critical difference from conventional hand-crafted feature representations. On the other hand, there are many parameters and hyper-parameters that affect the network's performance significantly. However there are not any systematic way to select the optimal value of these parameters and hyper-parameters. Therefore in the optimization module of the proposed method, we used BA to find the optimal architecture of the CNN. The proposed method was evaluated on a set of X-ray images collected in Shahid Beheshti University and the obtained results showed the excellent performance of proposed method.

Keyword: Dental caries, Diagnosis, Convolutional neural networks, Bee's algorithm, X-ray images.

PDF [ 674.26 Kb ] | Endnote File | XML