Estimation of surface topography for dental implants using advanced metrological technology and digital image processing techniques

摘要

Determination of surface topography for dental implants is crucial and possible with the advances of novel technological developments in surface measurement. Using these high technologies improves the production of the non-technical medical products and thus enhances the quality assurance for the surface of dental implants. The ultimate aim of this research study is to establish a ground in evaluating the surface topographical measurements to be done in real time. The current purpose of this study is to create a computerized model of a molar tooth and to investigate the surface characteristics of the original tooth and the implant samples not only in measuring their surface roughness but also in analyzing their surfaces using image processing techniques. The surfaces of dental implants and teeth were measured using 3D digital microscope and the stylus profilometer. The captured high resolution images were analyzed by Fast Fourier Transform (FFT) and the results from 3D digital microscope, the stylus profilometer and image analyses were compared. The measurement results showed that the surface roughness of the implant and tooth were significantly different. On the other hand, the correlation between the results from the measured values obtained from 3D digital microscope and the stylus profilometer, and those from image processing were very similar. The error between measured and predicted values from 3D digital microscope and image processing was calculated, respectively. The root mean square error (RMSE) is 0.033 μm for original tooth sample and 0.045 μm for implant sample. It can be concluded that the image processing methods can be used in roughness measurement process. Additionally, the image analysis results indicated that FFT analysis can represent the variation in the surface roughness with high correlation (R~2 = 0.94 for the original tooth sample and 0.88 for the implant sample).