Detection of Carious Lesions and Restorations Using Particle Swarm Optimization Algorithm

Autor: Mohammad Naebi, Eshaghali Saberi, Sirous Risbaf Fakour, Ahmad Naebi, Somayeh Hosseini Tabatabaei, Somayeh Ansari Moghadam, Elham Bozorgmehr, Nasim Davtalab Behnam, Hamidreza Azimi
Jazyk: angličtina
Rok vydání: 2016
Předmět:
Zdroj: International Journal of Dentistry, Vol 2016 (2016)
Druh dokumentu: article
ISSN: 1687-8728
1687-8736
DOI: 10.1155/2016/3264545
Popis: Background/Purpose. In terms of the detection of tooth diagnosis, no intelligent detection has been done up till now. Dentists just look at images and then they can detect the diagnosis position in tooth based on their experiences. Using new technologies, scientists will implement detection and repair of tooth diagnosis intelligently. In this paper, we have introduced one intelligent method for detection using particle swarm optimization (PSO) and our mathematical formulation. This method was applied to 2D special images. Using developing of our method, we can detect tooth diagnosis for all of 2D and 3D images. Materials and Methods. In recent years, it is possible to implement intelligent processing of images by high efficiency optimization algorithms in many applications especially for detection of dental caries and restoration without human intervention. In the present work, we explain PSO algorithm with our detection formula for detection of dental caries and restoration. Also image processing helped us to implement our method. And to do so, pictures taken by digital radiography systems of tooth are used. Results and Conclusion. We implement some mathematics formula for fitness of PSO. Our results show that this method can detect dental caries and restoration in digital radiography pictures with the good convergence. In fact, the error rate of this method was 8%, so that it can be implemented for detection of dental caries and restoration. Using some parameters, it is possible that the error rate can be even reduced below 0.5%.
Databáze: Directory of Open Access Journals
Externí odkaz: