Texture analysis of cone beam computed tomography images reveals dental implant stability

Abstract

The aim of this study was to characterize the alveolar bone of edentulous maxillary sites using texture analysis (TA) of cone beam computed tomography (CBCT) images and to correlate the results to the insertion torque, thus verifying whether TA is a predictive tool of final implant treatment. This study was conducted on patients who had received single implants in the maxilla (46 implants) 1 year earlier and whose torque values were properly recorded. Three cross-sections of the sites were selected on CBCT scans. Two regions of interest (ROIs) corresponding to the implant bone site and peri-implant bone were also outlined, according to virtual planning. The CBCT scans were exported to MaZda software, where the two ROIs were delimited following the previously demarcated contours. Values for the co-occurrence matrix were calculated for TA. With regard to the insertion torque value, there was a direct correlation with the contrast of the peri-implant bone (P < 0.001) and an inverse correlation with the entropy of the implant bone site (P = 0.006). A greater contrast indicates a greater torque value for insertion of the implants, and there is a possible association with a lower entropy value of the implant–bone interface.