Comparative analysis of the influence of post-processing filters and the variation of the parameters of acquisition in the quantification of metal artifacts in CBCT images

• Grant number: 19/00495-6
• Support type: Regular Research Grants
• Duration: August 01, 2019 - July 31, 2021
• Field of knowledge: Health Sciences - Dentistry - Dental Radiology
• Principal researcher: Sérgio Lúcio Pereira de Castro Lopes
• Grantee: Sérgio Lúcio Pereira de Castro Lopes
• Home Institution: Instituto de Ciência e Tecnologia (ICT). Universidade Estadual Paulista (UNESP). Campus de São José dos Campos. São José dos Campos , SP, Brazil

Abstract

Dental implants, in the region of acquisition of CBCT scans, lead to the formation of artifacts, which may interfere with the visualization and evaluation of the adjacent bone, making it difficult to follow up. The objective of this study will be to evaluate and compare, through the quantification of artifacts generated by implants, the effectiveness of image postprocessing filters, in the reduction of artifacts, available in two softwares, a national, e-Vol DX (e -Vol DX, CDT, Brazil) and a South Korean, Ondemand3D® (CyberMed, Seoul, Republic of Korea). Concomitant to variations in acquisition parameters of these images. TCFC images of a phantom composed of a waxed mandible with implants installed, varying the installation region of these voxels size and phantom position inside the FOV. The quantification of the artifacts will be carried out with and without the application of the filters. The images will be visualized in the ImageJ software to select the cuts and quantification of artifacts in two stages, by one radiologist. The calibration of the examiner will be performed by the intraclass correlation coefficient (ICC). The data will be submitted to the Shapiro-Wilks test for normality evaluation. Subsequently, the comparison of the type of implant, the filters of each software used and the dimensions of the voxel will be compared by the T test (if the distribution is normal) or the Mann-Whitney Rank Sun test if the distribution is not normal. The position of the phanton in the FOV will be analyzed by test of variance. (AU)